From Few to Many

Prior to the reforms, a quarter of the Colombian population had health insurance. Subsidies failed to reach the poor, wh...

Author: Inter-American Development Bank

16 downloads 784 Views 3MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Prior to the reforms, a quarter of the Colombian population had health insurance. Subsidies failed to reach the poor, who were vulnerable to catastrophic financial consequences of illness. Yet by 2008, 85 percent of the population benefited from health insurance. From Few to Many describes the challenges and benefits of implementing social health reforms in a developing country, exploring health care financing, institutional reform, the effects of political will on health care, and more. The reforms have provided important lessons not only for continued reform in Colombia, but also for other nations facing similar challenges. * * * * “Among the efforts to achieve universal health insurance coverage in low- and middle-income countries, Colombia stands out both for the long interval of implementation (since 1993) and for the thoroughness with which the experience has been analyzed and evaluated. Everything a researcher or policymaker might want to know about the country’s progress, setbacks and adaptations to changing economic and political circumstances is here in one impressive volume.” Philip Musgrove Deputy Editor Health Affairs “Colombia is a researcher’s dream: interesting reforms, exceptionally good data, and an engaging academic and policy community. Yet, little is known about the country because very few publications target the international audience. This book bridges that gap in the case of health reform by underscoring one of the most impressive accomplishments in the developing world. Although the Colombian reform still has many challenges, the book is a tool kit for those interested in improving the efficiency and equity in the delivery of health services.” Mauricio Cárdenas Senior Fellow and Director, Latin America Initiative The Brookings Institution

978-1-59782-073-8

www.brookings.edu

From Few to Many: Ten Years of Health Insurance Expansion in Colombia

From Few to Many is the first comprehensive look at Colombia’s 1993 health system reforms. It describes the implementation of universal health insurance, including a subsidized system for the poor, and examines the impact of this and other reforms during a time when Colombia experienced crushing recession and internal conflict that displaced half a million people.

From Few to Many Ten Years of

Health Insurance Expansion in Colombia

Copyright © by the Inter-American Development Bank. All rights reserved. For more information visit our website: www.iadb.org/pub

HEALTH

Amanda L. Glassman María-Luisa Escobar Antonio Giuffrida Ursula Giedion Editors

Regional Trade 10-20-09final.indd 158

10/20/09 2:42:48 PM


Ten Years of Health Insurance Expansion in Colombia

Amanda L. Glassman María-Luisa Escobar Antonio Giuffrida Ursula Giedion Editors

Inter-American Development Bank The Brookings Institution


From Few to Many

Inter-American Development Bank 1300 New York Avenue, N.W. Washington, D.C. 20577 www.iadb.org Co-published by The Brookings Institution 1775 Massachusetts Avenue, N.W. Washington, D.C. 20036 www.brookings.edu Produced by the IDB Office of External Relations The views and opinions expressed in this publication are those of the authors and do not necessarily reflect the official position of the InterAmerican Development Bank. Cataloging-in-Publication data provided by the Inter-American Development Bank Felipe Herrera Library From few to many: ten years of health insurance expansion in Colombia / Amanda L. Glassman … [et al.], editors.

p. cm. Includes bibliographical references. ISBN: 978-1-59782-073-8

1. Health insurance—Colombia—Case studies. 2. Health care reform— Colombia. 3. Medical policy—Colombia. 4. National health services— Colombia. 5. Public health—Colombia. I. Glassman, Amanda L. II. Inter-American Development Bank. III. Brookings Institution. RA412.5.C6 F76 2009 368.382 F9252--dc22

LCCN: 2009930145


©Inter-American Development Bank, 2009. All rights reserved. No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by information storage or retrieval system, without permission from the IDB.

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . v List of Abbreviations . . . . . . . . . . . . . . . . . . . . vii Chapter 1 Colombia: After a Decade of Health System Reform. 1 Background and Context . . . . . . . . . . . . . . . . . 1 A Decade of Change . . . . . . . . . . . . . . . . . . . 6 Chapter 2 Institutions, Spending, Programs, and Public Health . . . . . . . . . . . . . . . . . . . . . 15 Background . . . . . . . . . . . . . . . . . . . . . . 16 Program Case Studies. . . . . . . . . . . . . . . . . . 31 Discussion . . . . . . . . . . . . . . . . . . . . . . . 39 Chapter 3 The Impact of Subsidized Health Insurance on Health Status and on Access to and Use of Health Services. 47 Background and Context . . . . . . . . . . . . . . . . Methods. . . . . . . . . . . . . . . . . . . . . . . . Results. . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . .

49 53 55 69 71

Chapter 4 Public Hospitals and Health Care Reform. . . . . 75 Hospital Services before the Reforms of 1993 . . . . . . . . First Phase of the Reform: 1993–2002 . . . . . . . . . . . Reorganization, Modernization, and Redesign of the Public Hospital Networks: 2002 to Date. . . . . . . . . Conclusions and Lessons for Other Countries. . . . . . . .

76 80 88 95


Contents

Chapter 5 Financial Protection of Health Insurance. . . . . 103 The Colombian Health System . . . . . . . . . . . . . . Previous Research on Catastrophic and Impoverishing Health Expenditures in Colombia. . . . . . . . . . . Conceptual Framework . . . . . . . . . . . . . . . . . Data and Methodology. . . . . . . . . . . . . . . . . Descriptive Analysis . . . . . . . . . . . . . . . . . . Impact of Health Insurance on Financial Protection. . . . Conclusion . . . . . . . . . . . . . . . . . . . . . .

104 105 111 119 121 136 151

Chapter 6 Ten Years of Health System Reform: Health Care Financing Lessons from Colombia . . . . . . 157 Before the Reforms. . . . . . . . . . . . . . . . . . . Health Reforms of 1993. . . . . . . . . . . . . . . . . Results of the Reforms. . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . .

157 161 163 168 178

Contributors. . . . . . . . . . . . . . . . . . . . . . . 187


iv Contents

After more than a decade of implementation, health insurance in Colombia—once characterized as “managed competition in the tropics” by the late health minister and prominent economist Juan Luis Londoño—has achieved dramatic results in access, utilization, and financial protection, particularly for the poor. More than 85 percent of Colombia’s population is now insured. In the context of worldwide debates on the best way to achieve universal coverage without creating perverse incentives, this book brings empirically based analysis of Colombia’s achievements to audiences inside and outside of the country. The book also identifies challenges for the future in the areas of financing, public health, benefits packages, and hospital management, recognizing that health system reform is an ongoing process that requires continuous evaluation, learning, and adjustment. This book is the joint production of researchers based in Colombia and at the Brookings Institution and the Inter-American Development Bank. Initial work on the volume was supported by a grant from the Bill & Melinda Gates Foundation to the Brookings Institution, while additional research, editing, and publication costs were covered by the Inter-American Development Bank. These contributions are much appreciated. Kei Kawabata Manager, Social Sectors


Preface


This book is dedicated to the memory of Dr. Juan Luis Londoño, the visionary policymaker who set the Colombia reform in motion.

AIDS BCG CASEN DANE DHS DOTS DPT ECLAC EPS FEDESARROLLO FOSYGA GDP HIV LSMS MDD MPS NMCP

acquired immunodeficiency syndrome Bacillus Calmette-Guérin Encuesta de Caracterización Socioeconómica Nacional Departamento Administrativo Nacional de Estadística (National Administrative Statistics Department) Demographic and Health Survey directly observed treatment short-course diphtheria, pertussis, tetanus Economic Commission for Latin America and the Caribbean (Comisión Económica para América Latina y el Caribe) Entidades Promotoras de Salud (Health Promotion Entities) Fundación para la Educación Superior y el Desarrollo (Foundation for Higher Education and Development ) Fondo de Solidaridad y Garantía (Solidarity and Guarantee Fund) gross domestic product human immunodeficiency virus Living Standards Measurement Survey matched double difference Ministerio de la Protección Social (Ministry of Social Protection) National Malaria Control Program


List of Abbreviations

of Abbreviations

OECD PAB PAHO PBS POS PSM RDA SGSSS SISBEN SNS UNFPA WHO

Organisation for Economic Co-operation and Development Plan de Atención Básica (Basic Services Plan), now the Plan Básico de Salud Pan American Health Organization Plan Básico de Salud (Basic Services Plan) Plan Obligatorio de Salud (Compulsory Health Plan) propensity score matching regression discontinuity approach Sistema General de Seguridad Social en Salud (General System of Social Security in Health) Sistema de Identificación de Beneficiarios (Beneficiary Identification System) Sistema Nacional de Salud (National Health System) United Nations Fund for Population Activities World Health Organization


viii List

Colombia: After a Decade of Health System Reform María-Luisa Escobar, Ursula Giedion, Antonio Giuffrida, and Amanda L. Glassman

Background and Context Colombia is a middle-income country with an estimated 2005 population of 43 million (Departamento Administrativo Nacional de Estadística/National Administrative Statistics Department, 2007). Over the past three decades, the Colombian population has experienced the demographic and epidemiological changes that characterize societies in transition: a rapid decline in the total fertility rate (from 3.24 children per woman in 1985 to 2.48 in 2005), a significant increase in life expectancy (from 71.5 to 76.3 years for women and from 64.7 to 69 years for men, over the 1985 to 2005 period), and rapid urbanization (74.3 percent of the population lived in urban centers in 2005, compared with 67 percent in 1985). Half the population is identified as poor and inequality is widespread. Colombia, like other developing nations, is highly vulnerable to external and internal shocks that affect the income of the poor and their capacity to purchase needed health care services. Prior to 1993, only a quarter of Colombians had health insurance and more than half of total spending on health was out of pocket. Economic barriers were frequently cited as obstacles to care-seeking by the poor: nearly 60 percent of those who reported an illness requiring a visit to a health


chapter 1

and Glassman

facility in 1993 did not use these services because of costs associated with care-seeking. Colombia introduced mandatory social health insurance with the approval of an ambitious health care reform package in 1993. Occurring in the midst of decentralization and other state modernization reforms, the health reform was intended to increase burden-sharing of health risks and financing to improve access to care and provide financial protection to those beyond the formally employed. The reform introduced competition into both insurance and the provision of care through a managed-care model. As of 2008, more than 85 percent of the population is insured and access to and use of health care has increased significantly for the poor. Financial protection has also improved dramatically, as has spending on public health. Yet despite its novelty and promising results, the Colombian reform remains little studied or discussed internationally. Much of the extensive and high-quality literature produced in the country is not easily available to the rest of the world; perhaps this is one of the reasons little is known of the impact and challenges of Colombia’s introduction and implementation of health care reform. The experience offers an opportunity to understand the challenges, benefits, and pitfalls of introducing health system features like active purchasing, risk adjustment, insurance, and benefits packages—more common to wealthy countries—into a more resource- and capacityconstrained environment. This book aims to make recent research results public and to trigger an evidence-based discussion of this comprehensive reform, both nationally and internationally. The Health Care System before 1993 Prior to the changes introduced by the health care and financial decentralization reforms in 1993, access to and use of health care was low. The poor were vulnerable to impoverishing spending as a consequence of illness (Giedion, López, and Riveros, 2005). The health care system in Colombia was characterized by atomized risk pools, low efficiency, failure of public subsidies to reach the poor, large out-of-pocket expenditures, and significant inequality.


2 Escobar, Giedion, Giuffrida,

These factors disproportionately affected the poor: more than half of the bottom income quintile was unable to obtain care when they needed it because they could not afford it. One-quarter of the total population had no access to effective health care because of inadequacies in health care infrastructure, human resources, medicine, and medical goods (Barón, 2007). Although public facilities were intended to be free and were meant to cover the poor and uninsured, only 20 percent of individuals admitted to public hospitals were from the poorest income quintile and 91 percent of the poorest hospitalized patients incurred out-of-pocket expenses. Public subsidies benefited patients who were better off: almost 60 percent of admittances to public hospitals were of middle- or high-income individuals from the fourth and fifth income quintiles, but only 69 percent of the wealthiest hospitalized patients paid out-of-pocket expenses (Molina et al. 1993). The pre-reform National Health System comprised three independent sub-sectors: the official or public sector (government-owned facilities), the social security sector for formally employed people, and the private sector, used by both the insured and the uninsured. More than 40 percent of all health interventions and hospitalizations were provided through the private sector (Departamento Administrativo Nacional de Estadística, 1992). The system relied on general tax revenue, payroll contributions, and out-of-pocket expenditures, with no pooling of the three sources of financing. Not only was government spending before the reform low, but there was also no effective targeting mechanism for public subsidies. Colombia spent 1.4 percent of its gross domestic product (GDP) on health care (Molina et al., 1993) in 1993, though Mexico, Chile, Venezuela, Brazil, and Argentina were already spending a larger percentage of their GDP on health five years earlier. Public health financing was funneled to finance public hospitals, primary care facilities, public health programs, disease surveillance activities, and the administrative expenses of the central and decentralized Ministry of Health offices based on their historical budgets, without relationship to the level of services provided, the population’s health needs, or health outcomes. Beyond the centralized public health programs, there was no separate allocation of resources for disease prevention, health promotion, or community health activities. The public hospital network was composed of institutions of varied levels

3


Colombia: After a Decade of Health System Reform

and Glassman

of quality and efficiency but all with expensive labor costs stemming from a highly unionized workforce. The concurrent implementation of decentralization gave ownership of public facilities to local governments, which received National Treasury transfers to finance their historical budgets. There were few incentives for public hospitals to become more efficient, improve the quality of care, or adjust their portfolios of services according to population needs. In fact, many public hospitals were often in financial crisis by mid-year and relied on government bailouts to survive. People who were formally employed contributed with payroll taxes to social security institutions that provided health coverage to the enrolled population through their vertically integrated networks of facilities and health care providers. Social security beneficiaries represented around one-quarter of the Colombian population. Per capita health spending in the social security sector was several times higher than that for the rest of the population relying on the services of the Ministry of Health. In addition, a large private sector provided insurance products and health care to the population; insurance did not generally cover dependents. The Reforms of 1993 Law 100 of 1993 set up the legal framework of the new Colombian health care system and adopted the “structured pluralism” model (Londoño and Frenk, 1997). The reform unified the social security, public, and private sub-systems under the General System of Social Security in Health (known by its Spanish acronym, SGSSS). The reform also reorganized the system around functions and responsibilities rather than population groups. The 1993 health reform created mandatory universal health insurance to improve the equity and performance of public spending on health. Financed through a combination of payroll contributions and general taxation, this comprehensive national social insurance scheme included a contributory regime for those able to pay and a fully subsidized scheme for the poor. Beneficiaries enroll with public or private insurers (health funds), have legal rights to an explicit package of health benefits, and receive care from a mix of public and private



providers. The reform introduced a national equalization fund, the Fondo de Solidaridad y Garantía (FOSYGA; Solidarity and Guarantee Fund), to provide cross-subsidies between wealthy and poor, sick and healthy, old and young, and financing to stabilize health financing during economic crises. Both formally employed and independent workers earning more than a pre-determined minimum income must enroll in the contributory health insurance regime and contribute 12.5 percent of their income (12 percent, before January 2008). Funds are collected by the enrollee’s insurer of choice and then go to the national equalization fund. Poor and indigent people, who are identified as such through the Sistema de Identificación de Beneficiarios (SISBEN; Beneficiary Identification System), a proxy means test, do not make any insurance contributions and are covered under the subsidized health insurance regime. Insured individuals in both the contributory and subsidized regimes choose their insurer, choose care providers within the insurer’s network, and receive a health benefits package purchased by insurers from public and private providers through contracts. All participants in the contributory regime can enroll their dependents as a family unit. The benefits plan for the contributory regime is generous and covers all levels of care. The package had a premium equivalent to US$207 annually in 2007. Primary care, some inpatient care, and emergency care are now covered under the subsidized regime and have a premium equivalent to US$117. This coverage is complemented by inpatient care at level 3 public hospitals. According to the law, the supply-side subsidies should gradually transform into demand-side subsidies as insurance coverage expands, eventually leading to universal coverage with a uniform package for everyone. Residents still uninsured are able to use public facilities to receive preventive and public health services and emergency care. Regardless of insurance status, all citizens are eligible to receive the benefits of the public health intervention package, the Plan Básico de Salud (PBS or Basic Services Plan; called the Plan de Atención Básica until 2008). Municipalities provide health promotion and disease prevention services included in the PBS. Financing for public health is separate from other health care funding.

5



and Glassman

The reforms mandated that public hospitals would make the transition from being state care providers financed through supply-side subsidies based on their historical budgets, to being state enterprises with autonomous governance structures remunerated for the services provided. Private health care providers were to compete with public providers for the provision of the mandatory benefit plan on the basis of quality and were to negotiate contracts with insurers. The challenges were many and the pressure for modernization in the public hospital network was great with the changes introduced to the provision of care.

A Decade of Change The Political Economy The government administration changed with the presidential elections in mid-1994, seven months after Law 100 was approved. Although from the same political party as the previous government, the new team was not completely aligned with the principles of the reform. Approval of key by-laws and regulations required for implementation of the law were delayed and the reform process lost momentum. Despite these difficulties, however, the contributory regime attracted new insurers that entered the system to extend insurance coverage. Regulations for insurers for the subsidized regime were formally introduced at the end of 1995 to launch the implementation of that scheme. Political difficulties and necessary negotiations with local governments followed; the subsidized regime was not launched until almost two years later. Between 1991 and 1994, Colombia experienced important economic growth, followed by a dramatic reverse that led to a recession in 1998–99 (with record negative growth of −4.3 percent in GDP in 1999). A mild economic recovery followed in 2000–01, with GDP growing in those years by 2.8 and 2.4 percent, respectively. Official unemployment figures rose from 8.7 percent in 1995 to 20.2 percent in 2000, however (representing the highest unemployment rates in the past 20 years), and in 2000, informal employment represented 54.9 percent of total employment. The recession occurred in the context of an intensification of the internal armed conflict, which displaced about 580,000 people between 1998 and 2001. The rural population was the most



severely affected: 82 percent of displaced individuals came to cities from rural areas. The health care reforms had been only partially implemented by 2001 and the SGSSS was undergoing a severe and generalized financial crisis. Universal insurance coverage was still far from being achieved in 2001, with only 58 percent of the population insured, and the transformation of hospital financing had affected only 50 percent of hospital revenue. Confusion about the decentralized roles of local authorities in public health, combined with shortages in the allocation of resources for vaccination programs, negatively affected immunization rates. That situation forced the government to consider two alternatives. One was to return to the supply-side subsidies, with public-sector budgets controlled by the central government—and in particular the National Treasury—but at the expense of the subsidized regime and the health care system’s reform (Gaviria, Medina, and Mejía, 2006). Alternatively, the government could correct the external conditions affecting the delivery of care and strengthen the health sector reform process. The government adopted the latter approach and the administration committed to accelerating the expansion of subsidized health insurance for the poor; developing a program to support the redesign, reorganization, and modernization of public hospitals and to ensure their financial sustainability; and strengthening the national immunization program. The implementation of this vision began in 2002. The previous labor and health ministries were merged. The new Ministry of Social Protection became responsible for pensions, health insurance, public health programs, and all other social assistance programs. A quality assurance system was designed, with the introduction of a licensing and accreditation process for public and private health care facilities and providers. An aggressive hospital restructuring program was negotiated with local governments and the Ministry of Finance. Measuring Results To objectively measure the impact of social policy change in the developing world, it is necessary to analyze progress in light of the original pre-reform conditions, not only with respect to the degree of achieve-

7



and Glassman

ment of ambitious reform goals. Given that reforms are processes evolving over time and within societies in states of continuous change, it seems sensible to first understand the complexities of transformation in order to objectively assess any change, even when it seems small and incomplete by international standards. Breaking apart the traditional social security schemes for the formally employed and transforming them into regular, competing insurers was a political and institutional task impossible to imagine before 1993. In fact, most—if not all—countries in Latin America with health care systems similar to that of pre-reform Colombia still have segmented health care systems with significant inequality in health financing, no explicit benefits packages, and no contracting of a mix of public and private providers. Establishing a functional equalization fund to transform income contributions into risk-adjusted capitated payments to insurers was a test for those financial agencies to be contracted through public bidding to manage the fund’s finances. The complexity of the equalization fund—with four sub-funds (or accounts in FOSYGA) to support such functions as full or partial insurance premiums for more than 30 million people—requires well-developed capital and financial markets accompanied by stateof-the-art information systems. Demonstrating and accepting that public subsidies did not reach the poor, and introducing a proxy means test to better target government subsidies to those most in need, was an immense challenge in the early 1990s; it still is in many parts of the developing world. The introduction and use of the SISBEN in the health sector was a victory for the Colombian poor and an important improvement for the allocation of public resources to health. The scheme was later adopted in other sectors as well. Governance mechanisms like the Consejo Nacional de Seguridad Social (National Social Security Council)—with representatives from public and private insurers and care providers, the government, and civil society having the power to make decisions on the functioning of the health care system—are still unknown in many countries with income levels similar to Colombia’s. After 1993, for the first time there is a formal regulatory structure, through which the Minister of Finance and the Minister of Health sit at the same table to debate the techni-



cal and financial aspects of the health care system when negotiating any decision affecting public finances. An open negotiating sphere in which all special interest groups are represented is commoner to more egalitarian societies with well-established democracy than to a low- to middle-income country with a 40-year history of internal armed conflict. The risk of capture was important and the technical requirements for it to function as envisioned were great. Ten years of implementation have taught important lessons both for Colombia and for other countries that face similar challenges. The five papers brought together in this volume examine Colombia’s health system reforms and their impact after more than a decade of implementation. The book presents discussion in areas such as financing, hospital reform, insurance impact, regulation, and public health. Each paper analyzes the reform from a different perspective, although all are naturally inter-related, given the structure of the system and the way it functions. The analysis discussed here refers to the period between 1993 and 2003; it was carried out with the information available before the most recently released National Health Survey of 2007–08 and the approval of Health Law 1122 in 2007. Examination of the Reform Experience Chapter 2, by Amanda L. Glassman, Diana M. Pinto, Leslie F. Stone, and Juan Gonzalo López, seeks to improve the quality of the policy debate on public health in Colombia by examining the evolution of public health institutions, spending, and programs—and the effectiveness of these—over the past 30 years. The chapter uses the vaccination, tuberculosis, and malaria prevention and control programs as case studies. The authors find that public health conditions have improved substantially in Colombia over the past decade. Equity in access to public health services has increased over time, but remains a problem for the very poor and for ethnic minorities and displaced people. Spending on public health has increased, and earmarked financing protects it in the aggregate. A severe recession in the late 1990s negatively affected the availability of non-earmarked financing for public health, however, which led to drops in health coverage during this period. Insurance has proven a useful tool to increase coverage rates for some interventions,

9



and Glassman

although available data and analyses provide a confusing picture of coverage and impact trends in tuberculosis and malaria. Decentralization reforms have complicated the public health panorama, particularly from the perspective of vulnerable populations, leading to suboptimal implementation of programs and, perhaps, outcomes. The use of insurance and contracting to achieve public health goals is of interest worldwide, and the Colombia case shows that the devil is in the details of underlying governance, data, and evidence necessary to develop and implement effective policy. Chapter 3, by Ursula Giedion, Beatriz Yadira Díaz, Eduardo Andrés Alfonso, and William D. Savedoff, examines the impact of health insurance by applying a series of different quasi-experimental design techniques, including regression discontinuity, propensity score matching, and matched double difference when comparing differences between insured and uninsured people. The chapter discusses the effect of subsidized insurance on equity, access to care, utilization of services, and financial protection of households. Although insurance coverage increased across all income groups after 1993, the improvement has been particularly pronounced among the poorest individuals and in the least-developed regions. Empirical evidence indicates that before the reforms, the poorest segment of the population had almost no financial protection when facing illness, since only a small portion of costs were covered by health insurance. Meanwhile, 6 of every 10 of the wealthiest individuals were protected by insurance. A decade later, the gap between the rich and the poor has been reduced considerably. Insurance coverage in the lowest income group has increased to 18 times what it was in 1993, whereas coverage among the highest income group increased only 1.4 times. Analysis with four methodologies consistently indicates that the subsidized health insurance scheme has considerably improved access to and utilization of health services, especially among rural and poor Colombians. Insured people of all ages are much more likely than their uninsured peers to receive care when they need it. Analysis results show that insurance is quite important for rural and poor children because it increases the likelihood of prenatal care, of attendance by a qualified care provider at birth, of receiving care when ill, and of a completed immunization scheme.



Chapter 4, by Teresa M. Tono, Enriqueta Cueto, Antonio Giuffrida, Carlos H. Arango, and Alvaro López, presents evidence and discussion of the transformation of the public hospital network and of the achievements, failures, difficulties, and challenges the health care system still faces. Although the reform laws gave public hospitals the legal framework to become more autonomous entities, hospitals had no precedent for operating in a competitive environment, and had high labor costs and few managerial skills. The latter problems were great challenges for public hospitals to overcome on their own. In response, a modernization project tailored to the shortcomings of each individual hospital was set in place to improve both the capacity of public hospitals to participate in the health services delivery market, and their productivity and the quality of services they offered. By 2006, 179 public hospitals had already participated in this ongoing process, some with good results. The hospital modernization experience shows that public hospitals were not able to modernize on their own, even though an appropriate legal framework was in place. Maintaining strong political will over time is necessary for successful transformation of public facilities. Skillful negotiation with decentralized governments has also been necessary to provide appropriate incentives to develop a lasting process of transformation. An appropriate allocation of resources is also required, making reshaping of the public hospital network costly and slow. The results presented here suggest that legislation, along with hospital network modernization and labor restructuring programs, improves the efficiency and quality of the hospitals: participating public hospitals have decreased their deficits and improved their market participation. Chapter 5, by Carmen Elisa Flórez, Ursula Giedion, Renata Pardo, and Eduardo Andrés Alfonso, analyzes the impact of the reforms on financial protection of health insurance. This chapter discusses the methodological challenges of measuring financial protection and the sensitivity of results to the method used. Results show that the reforms provide substantial financial protection from catastrophic expenditure and impoverishment, benefiting all insured people in both the subsidized and contributory regimes, particularly self-employed and informally employed workers.

11



and Glassman

Finally, Chapter 6, by María-Luisa Escobar, Ursula Giedion, Olga Lucía Acosta, Ramón Castaño, Diana M. Pinto, and Fernando Ruiz Gómez, presents evidence of the impact of the reforms on the level, composition, distribution, and equity of health care financing. The chapter also examines threats to the reform’s financial sustainability. The health care system is still financed by both general tax revenue and payroll contributions; however, its financial structure and the mechanics of resource flows were changed to improve equity, to extend insurance coverage to all—the poor in particular—and to improve efficiency of public spending. The composition of financing in Colombia is now similar to that of countries that are part of the Organisation for Economic Co-operation and Development (OECD); public spending, including social security, accounts for more than 80 percent of total health spending, while outof-pocket spending is among the lowest in the world. Results support the idea that the reforms make government subsidies for health the besttargeted public subsidy in the country. The subsidies have also had an important redistributive impact. Despite these major accomplishments, the system faces important challenges before it can achieve financially sustainable universal coverage. Despite these encouraging results, there is still much to do and to improve. A decade after the reform, 15 percent of the population remains uninsured; benefit plans under the contributory regime and the subsidized regime still differ. There are deficiencies in the quality of care and not all public hospitals are modernized. The stewardship function needs to be strengthened; the financial sustainability of the system is continually at risk. Nevertheless, the health care system in Colombia experienced drastic changes that have benefited the health of the country’s population.



References Barón, G., ed. 2007. Cuentas de salud de Colombia 1993–2003. El gasto de salud y su financiamiento. Bogotá: Ministerio de la Protección Social. Departamento Administrativo Nacional de Estadística. 1992. Encuesta Nacional de Hogares [National Household Survey]. Bogotá: DANE. ———. 2007. Proyecciones nacionales y departamentales de población 2006–2020. Bogotá: DANE. Gaviria, A., C. Medina, and C. Mejía. 2006. Evaluating the Impact of Health Care Reform in Colombia: From Theory to Practice. Center for Economic and Development Studies Document No. 6. Bogotá: Universidad de los Andes. Giedion, U., A. López, and H. Riveros. 2005. Opciones para la transformación de subsidios de oferta a demanda. Washington: InterAmerican Development Bank. Londoño J.L., and J. Frenk. 1997. Structured Pluralism: Towards an Innovative Model for Health System Reform in Latin America. Health Policy 41:1–36. Molina, C.G., M.C. Rueda, M. Alviar, et al. 1993. Estudio de incidencia del gasto público social: el gasto público en salud y distribución de subsidios en Colombia. Bogotá: World Bank, FEDESARROLLO.

13




Institutions, Spending, Programs, and Public Health Amanda L. Glassman, Diana M. Pinto, Leslie F. Stone, and Juan Gonzalo López

D

uring a decade of health insurance and decentralization reforms, and despite a profound economic recession in the late 1990s and an ongoing internal armed conflict that has waxed and waned, average indicators of health and well-being have improved substantially in Colombia (Table 2.1). For example, the infant mortality rate in 2005 was lower than that in Brazil (28 per 1,000) and Mexico (22 per 1,000), two comparable middle-income countries in the region (WHO, 2007). Yet nowhere have Colombia’s reforms been as controversial as in their impact on public health. Much of the literature reaches conclusions about the impact of the insurance and decentralization reforms based on limited data and inappropriate methods of analysis. An article examining the evolution of vaccination coverage in the late 1990s, published by the Pan American Health Organization (PAHO), for example, concludes that “[p]ublic health programs in Colombia have deteriorated…. Health systems based on regulated competition are not the most suitable ones for Latin America” (Homedes and Ugalde, 2005). In a news item published by the British Medical Journal in 1997, the correspondent concludes that the health status in Bogotá is worsening owing to the reforms, which have forced physicians to work longer hours (Richards, 1997). This chapter seeks to improve the quality of the policy debate on public health in Colombia by examining the evolution of public health


chapter 2

and López

Table 2.1 Public Health and Living Standards Before and After the 1993 Reforms Indicator

ca. 1990

ca. 2006

Public spending on public health, excluding donor funding and supply subsidies (billions of 2004 Colombian pesos)

No data

1,417,000

Public spending on public health (percentage of GDP)

No data

0.41

Unmet basic needs, such as clean water, sewage, etc. (percentage of total population with at least one basic need unmet)

35.8

27.6

Life expectancy at birth (years)

68.3

72.8

Infant mortality rate (per 1,000 live births)

26.3

17.2

Under-5 mortality rate (per 1,000 population)

34.7

21.4

Births attended by professionals (percentage of total births)

81.8

96.4

Measles, mumps, and rubella immunization (percentage of children aged 12–23 months)

82.0

89.0

Sources: Spending: authors’ analysis based on Ministry of Social Protection data; basic needs and life expectancy: National Administrative Statistics Department (www.dane.gov.co, accessed June 30, 2007); remainder: Demographic Health Survey 2005; Profamilia and Macro International (2006).

institutions, spending, and programs—and their effectiveness—over the past 30 years. Vaccination, tuberculosis, and malaria prevention and control programs are used as case studies. It is hoped that this synthesis and discussion of the evidence, developed using available data and literature, will be relevant both inside and outside Colombia, as other countries grapple with similar policy issues worldwide. The chapter will first provide background on the health and decentralization reforms and their impact on financing and spending for public health, as well as trends in the burden of disease and mortality statistics. This background sets the stage for the examination of the three program case studies, followed by a discussion.

Background Health Reforms Prior to 1985, public health interventions—defined as maternal and child care, and control of epidemics, and later including immuniza-


16 Glassman, Pinto, Stone,

tion and control of tuberculosis, leprosy, malaria, and sexually transmitted diseases—were financed as an indistinguishable part of the then–Ministry of Health budget and were organized as centralized programs. Seventy-two percent of health care financing was spent on curative care services that disproportionately benefited relatively well-off patients (analysis based on National Health Accounts data produced by the National Planning Department). Public financing for public health programs such as immunization and family planning was complemented by international donors such as the United States Agency for International Development (USAID), the United Nations Fund for Population Activities (UNFPA), the World Health Organization (WHO), and PAHO (Tono et al., 2002), although the exact amounts of these contributions are impossible to quantify retrospectively. Public health interventions were also provided by the Social Security Institute, a social insurance scheme for formally employed workers (but excluding their dependents), financed by a payroll tax. Facilities owned by the institute provided these services. After 1993, motivated by the poor performance of the health system and the high levels of out-of-pocket spending on health care, the financing and care provision arrangements governing public health were substantially changed. National health insurance covering formally employed workers and their families, and progressively larger numbers of the poor, was introduced with Law 100 of 1993. The insurance scheme—intended to be universal eventually—was made up of a contributory regime of formal sector workers and their families, and a fully subsidized regime directed to the poor. (Legislation affecting public health is embodied in the original reform law—Law 100 of 1993—as well as in the law governing decentralization—Law 60 of 1993—and subsequently in Law 715 of 2001, which attempted to clarify public health functions and responsibilities at each level of government.) The law distinguished between a package of health interventions for individuals, known as the Plan Obligatorio de Salud (POS; Compulsory Health Plan), to be financed and purchased by private and public insurers, and a package of public health interventions, known as the Plan Básico de Salud (PBS; Basic Services Plan). Called the Plan de Atención Básica until 2008, the PBS was to be financed by a mix of public resources, and resources purchased and/or provided by sub-national

17


Institutions, Spending, Programs, and Public Health

and López

(departmental and municipal) governments. National health insurance for the poor—the subsidized regime—is financed by contributors to the formal sector contributory regime, as well as by general revenues and other earmarked taxes (see Chapter 6). Thus, insurance for the poor is also financed by public monies. In 2001, a law governing responsibilities at different levels of government required that departmental health directorates contract out for PBS activities. However, a later circular (No. 0018 of 2004) from the Ministry of Social Protection required that departments or municipalities contract preferentially with public providers: “…if quality conditions are equal, it will be preferable to contract with public rather than private providers.” Insurance-financed interventions are conceptually categorized as those interventions with benefits that accrue mostly to individuals, while PBS interventions are those with benefits that are collective or display high externalities. Insurance-financed interventions reach the insured person, while PBS interventions are intended to be universal. In addition to the PBS, sub-national governments are required to provide laboratory services and individual services for uninsured people during the transition to universal coverage. This supply-side subsidy for the uninsured has been ill-defined and left to the discretion of each municipality, however; many municipalities simply transferred the funds to public hospitals. Studies have found lower rates of utilization and coverage of key interventions among uninsured patients, suggesting that subsidies channeled to public hospitals for this purpose are not being optimally used (see Chapter 4). Both sets of interventions, along with protocols and standards of care as of 2000, were explicitly established in laws, norms, and guidelines, thus creating a financing and expenditure benchmark for public health and a legal entitlement for the respective target populations.1 Table 2.2 describes the interventions, target populations, and financing sources for public health in Colombia in 2006. Some overlap in the content of packages exists, particularly in chronic disease control.

1 Colombia’s Constitution of 1991 allows easy access to the court system; Colombians are able to, and frequently do, contest health access problems. See Chapter 6.



Table 2.2 Public Health Service Packages, Target Populations, and Financing Sources Target population

Financing sources and amount

Basic Services Plan • Law 100/1993 and Resolution 4288/1996: Public information; education; health promotion; control of tobacco, alcohol, and drug abuse; nutritional supplementation; family planning; deparasitization; vector control; environmental, food, and animal safety; national campaigns for prevention, early diagnosis, and control of contagious diseases such as HIV/AIDS, sexually transmitted diseases, tuberculosis, and leprosy, and tropical diseases such as malaria • Circulars 018 and 002/2004: Priority chronic disease risk-factor screening

Universal (see coverage rates by intervention, in next section)

• Source: National transfers allocated to public health • Amount: 10.4% of total national transfers = 399 billion 2006 pesos

Compulsory Health Plan for the Subsidized and Contributory Insurance Regimes—public health content • Resolution 3997/1996: Prevention of diseases related to pregnancy, birth, and puerperium; child growth monitoring; child vision and hearing disease prevention; acute respiratory infection prevention; immunization; drug addiction prevention; cancer and other chronic disease prevention • Agreement 117/1998: Pregnancy, birth, newborn, and low-birth-weight interventions; integrated management of childhood illnesses; preventive oral care; priority chronic disease riskfactor screening and some treatment (hypertension, diabetes, obesity, asthma)

Insured Contributory regime: • Source: Wage contributions people: • Contributo- • Amount: Total resources ry regime: disbursed for premiuma = 5 trillion 2006 pesos. 15.9 Resources for promotion and million • Subsidized prevention sub-fund = regime: 235 billion 2006 pesos. 18.3 Subsidized regime: million • Source: National transfers for • 70% total demand subsidies, Solidarity population and Guarantee Fund (FOSYGA), sub-national resources • Amount: Total resources disbursed to cover subsidized regime premium = 3.8 trillion 2006 pesos. 4.01% transferred to municipalities for promotion and prevention = 157 billion 2006 pesosb

Package name and content

Source: National Planning Department. a Premium covers full benefits package for each regime. Insurers are expected to spend at least 10% of premium in health promotion and disease preventive activities. b Transfer was eliminated by Law 1122 of 2007.

19



and López

Decentralization Reforms In the mid-1980s, under pressure to democratize and decentralize government, Colombia began to implement fiscal, political, and institutional decentralization reforms that sought to reassign government functions and responsibilities among the national, departmental, and municipal levels.2 Under this decentralization framework, the central government’s role concentrates on policy design, regulation, and public finance. Departmental governments assume regional planning, management, and finance responsibilities, and provide some services and articulation of local and national levels. Municipal governments take on policy implementation and public service provision. From 1990 to 1993, legislative mandates introduced additional sub-national functions and responsibilities, and defined new sources of financing for health service provision and their respective allocation formulas. 3 Administrative procedures to certify local governments as “decentralized” were established. If met, these procedures shifted authority, responsibility, and budgetary control of these resources to departments and municipalities.4 Among these requirements was the creation of local health directorates that would assume public health responsibilities. Health policy and decentralization reforms thus combined to distribute public health responsibilities as shown in Table 2.3. Implementation of decentralization was heterogeneous in terms of the depth to which territories carried out the processes required to assume the public health functions established in the law—and the speed with which they did it. This problem has been attributed to lack of clarity and precision in the laws concerning responsibilities at the different levels of government, poor articulation between national health sector policies and the new functions that were to be assumed 2 Colombia is divided into 32 administrative units, or departments, which in turn are divided into municipalities, of which there are 1,098. There are also four capital districts corresponding to major cities. Municipalities are governed by mayors and departments by governors, both elected by popular vote. Although 70% of municipalities are rural and have fewer than 20,000 inhabitants, more than 60% of the population lives in the six largest, urban municipalities. 3 Law 10/1990; Political Constitution, 1991; Law 60/1993 and Decree 1757/1994; Law 100/1993. 4 Law 10/1990 Section 37; Law 60/1993 Sections 14 and 16; Decree 1770/1994.



Table 2.3

Public Health Responsibility by Level of Government Responsible entity

Public health function

Insurance carriers

Departmental Municipal health health directorates directorate

Individual/ family health promotion and disease prevention; curative care for conditions of public health interest

Provision of individual services to insured patients

Provision of individual services for uninsured patients (certified municipalities)

Provision of individual services for uninsured patients (non-certified municipalities)

Ministry of Health: Purchase and distribution of medications for tuberculosis, leprosy, leishmaniasis, and malaria, and supplies for public health laboratory diagnostic tests

Collective health promotion and disease prevention actions

Pre-2006: Purchase of syringes for vaccinating insured patients

Provision of collective care; hiring vaccinators and fumigators

Carrying out complementary municipal activities

Ministry of Social Protection: Complementary sub-national activities; acquisition and distribution of vaccines and supplies for implementation of the Programa Ampliado de Inmunización and vector control

Public health information and surveillance

—

Data collection and analysis for conditions of public health interest; case follow-up, outreach, and referral for diagnosis and treatment of contagious diseases; control of epidemics

—

National Health Institute: Planning, development, and articulation of sub-national surveillance system; design of standards; and provision of technical assistance

Central government

Continued on next page

21



Table 2.3

and López

Public Health Responsibility (continued) Responsible entity

Public health function

Insurance carriers

Departmental Municipal health health directorates directorate

Environmental risks

—

Oversight of water, food, disease vectors, and risk factors for infectious diseases

Community participation

—

Institutional capacity building

Central government

Control of medications and potentially toxic chemicals

Food and Drug Safety Agency (INVIMA): Training, assistance, and control of sub-national governments in implementation of norms and procedures relating to medications and chemical substances

Provision of information on health rights and responsibilities, promotion of community participation

—

—

—

—

Provision of technical assistance, supervision, and evaluation of municipal PBS

Research

—

—

—

—

Stewardship, planning, and monitoring

—

Established Consejo Territorial de Salud (Territorial Health Council)

Development of departmental PBS complementing municipal activities; distribution of resources for public health to non-certified municipalities

Ministry of Social Protection: Development of national policies and guidelines for PBS activities, inter-sectoral activities; National Health Supervisory Agency: Inspection and oversight of efficient use of public health resources

Provision of technical assistance; supervision and evaluation of subnational PBS




Table 2.3

Public Health Responsibility (continued) Responsible entity

Public health function Public health laboratories

Insurance carriers —

Departmental Municipal health health directorates directorate —

Provision of public health laboratory services

Central government National Health Institute: Coordination, assistance, and supervision of national network of public health labs

Source: Authors.

by territories, differences in sub-national financial and administrative capacity, fluctuations and lack of stability in available resources, and lack of surveillance and control over sub-national performance (Vargas and Sarmiento, 1997; Sánchez, Yepes, and Cantor, 1998; Sánchez and Yepes, 1999; Herrera and Cortez, 2000). Evaluations of departmental and municipal uptake of PBS interventions have focused on the number of municipalities taking on and assigning staff to the implementation of the PBS, the frequency of supervisory and technical assistance visits, and the application of norms and standards associated with the PBS and its contracting (Grupo de Gestión Integral en Salud, 2005; Jaramillo, 1999; Unión Temporal, 2004). Over time, an increasing number of municipalities have taken on the PBS and about half directly executed fund transfers for that use (prior to regulations in 2001 stipulating that there would be no direct provision). Contracting processes have worked well for most municipalities but have been problematic for about one-third: 36 percent reported difficulties, while a substantial portion did not follow minimum standards for due process (no evaluation of timeliness or quality, frequent resort to direct contracting without competitive bids, no supervision, or cancellation for non-performance) (Unión Temporal CCRP-ASSALUDBDO, 2004). Some municipalities used funds for purposes not permitted by legislation, such as for hiring personnel to work directly in the municipality, and a large proportion contracted with public hospitals

23



and López

for services such as environmental safety and disease vector control, in spite of the limited expertise and poor track record of these entities in this area (public hospitals are run by the municipality and were extensively used for patronage during the 1995–2000 period; see Chapter 4). Accounting problems and misuse of funds are also reported in some cases (Grupo de Gestión Integral en Salud, 2003–05). Epidemiological notification and management reporting systems are sluggish. Although most eventually report, 92 percent of municipalities did not comply with required reporting on time in 2003; a year later the figure had decreased to a still-high 85 percent. Departments were found to have been lax in their role of advising, monitoring, and enforcement, although the participation of government and civil society in the development of PBS action plans was high (Unión Temporal, 2004). Shortcomings observed have been attributed to high human resources turnover rates, poor skill mix, poor-quality information systems that generate incorrect or unreliable data, absence of effective civil society oversight mechanisms, late and ineffective interventions by controllers and auditors in response to complaints, and low population awareness of rights and responsibilities in public health. Governments have made efforts to align incentives better in the system and to assess the impact on public health of the new arrangements. Slow progress in meeting decentralization goals, and an increasingly precarious fiscal situation at the sub-national level prompted enactment of Law 715 in 2001. This law sought to correct the weaknesses identified in previous policy. Law 715 reset the amount of national fiscal resources for health and the parameters used for their distribution, basing the latter on sub-national indicators of equity and efficiency. The law also redefined responsibilities to be more in accordance with sub-national capacity. For example, less-developed municipalities are no longer responsible for vector control and environmental health. However, the law continues to permit decentralized municipalities to maintain functions and authority over resources and service provision, as long as they meet performance targets designed for this purpose. The law also increases the department’s role in articulation and oversight of public health activities carried out by insurers and municipalities.



Vaccination rates are among the performance targets established under the new laws; their achievement is rewarded with a bonus payment to municipalities. This system, however, while conceptually appealing as a pay-for-performance mechanism, created unintended effects through the use of official denominators from the 1993 census. Given extensive economic- and conflict-related migration after 1993, the reliability of municipal-level projections for disaggregated age groups (0–11 months, 12–24 months) across such a long period rapidly deteriorated, thus greatly distorting denominators used to calculate official vaccination rates. A 2004 sample survey carried out to check administrative data quality at the municipal level found errors that consistently both over- and under-reported vaccination performance by large margins, resulting in both undeserved rewards and performance improvements that went unrecognized. For example, in Quibdó, the capital of Chocó, the national statistical agency grossly underestimated the growth in the population of children under 1 year old (the denominator), thus resulting in an “official” diphtheria/pertussis/tetanus (DPT) vaccination rate of 111 percent for 2003. Results from the 2004 sample survey showed a DPT coverage rate of 49 percent for this city. Conversely, in Valparaiso Antioquia, as with many other small towns, the projections for the population of children under age 1 were overestimated, resulting in an official DPT vaccination rate of 63 percent, whereas the sample survey showed a coverage rate of 93 percent. Law 715 also introduced greater fragmentation of public health activities aimed at individuals included in the subsidized regime benefits package. It did this by shifting the provision of specific health promotion activities, immunization, family planning, and cervical cancer screening to municipalities.5 Under this law, for example, a beneficiary enrolled in the subsidized insurance regime would be sent to a municipally financed care provider for a Pap smear. If an abnormal smear required a confirmatory diagnostic test (colposcopy), the patient would have to pay for the test out of pocket because that intervention was not covered by the subsidized regime benefits package. If diagnosed with cancer,

5

Law 715/2001 Section 46; Agreement 229/2002.

25



and López

the patient would then return to the insurance-financed provider for treatment, which is covered under the benefits package. Municipalities had little incentive for screening and early diagnosis, since reimbursements for treatment accrued to insurers. The law has since changed again, but continuity of care for some priority interventions for both insured and uninsured populations remains problematic. Financing and Spending The main source of funding for health care before 1993 was central government revenue allocated to the Ministry of Health for specific programs or transferred to sub-national (departmental and municipal) governments. At the sub-national level, sources of funding included local taxes earmarked for health and other sources of revenue allocated at the discretion of each sub-national government. No data are available on the allocation of resources for public health prior to 1993 because budgets were transferred in lump sums, and expenditures on public health were made at the discretion of local health authorities. Between 1970 and 1990, the share of expenditures on personal care increased from 50 percent to 72 percent, while expenditures on environmental interventions and infrastructure decreased from 31 percent to 12 percent and from 21 percent to 16 percent, respectively.6 Although all these expenditure categories include interventions that could be considered part of the public health armament, environmental and infrastructure expenditures are likely to contain a larger share of public health expenses. For example, the bulk of personal care expenditures were distributed among hospital, medical, and dental care (about 55, 27, and 5 percent of the total, respectively); the remainder was allocated to interventions related to public health such as nursing, health promotion, and immunization services. 6 Values obtained for 1970–90 data on allocation of total public expenditures on health for three purposes: personal care (medical care and other services provided to individuals), environmental interventions (programs and interventions to reduce risk factors, such as aqueducts, sewage systems, vector control campaigns, food safety, etc.), and infrastructure; no other expenditures, such as capacity building, construction, research, and health promotion activities for children and elderly people, are included (calculations based on data in Molina et al., 1994, and Vivas et al., 1988).



The health and decentralization reforms not only increased resources for public health but also earmarked them exclusively for this purpose, either for collective interventions through PBS or for individual services provided through the insurance benefits packages. Given the stagnation in insurance coverage as a result of the recession, limited discretionary funding was also provided by municipalities to finance individual services for uninsured people, usually through transfers to public hospitals. In addition, a special sub-fund for health promotion and disease prevention activities for contributory regime enrollees was created, equivalent to 0.41 percent of total revenue from premium contributions. The growth and distribution of resources for public health, categorized by purpose, from 1995 to 2004 are shown in Figure 2.1. Public health resources underlying the calculations used in Figure 2.1 include resources from the national budget allocated to the Ministry of Social Protection, national transfers for public health (Situado fiscal 1995–2002, Sistema General de Participaciones 2002–04), a proportion of the contributive and subsidized premiums expected Figure 2.1 Resources for Public Health by Purpose, 1995–2004 (2004 millions of pesos) 1,500,000 1,400,000 1,300,000 1,200,000 1,100,000 1,000,000 900,000 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0

1995

Ministry programs & INS Individual services (POS-S) Individual services (CR-subfund) Collective interventions (PAB) Individual services (POS-C)

14%

35%

17%

25% 10%

1996

1997

1998

1999

2000

2001

2002

2003

2004

Source: Authors’ calculations based on Ministry of Social Protection data. INS = National Institute of Health; PBS = Basic Services Plan; POS-S = subsidized regime; POS-C and CR = contributory regime.

27



and López

to be assigned to health promotion and disease prevention activities (10 percent), and the promotion and prevention contributory regime sub-fund revenue. No data were available on external donor funding or supply subsidies for insured patients. Resources available for public health increased by 30 percent between 1995 and 2004. Total resources for public health in 2006, excluding donor funds, could have totaled $Col 1.4 trillion (US$584,0007), or 0.4 percent of the GDP. With respect to purposes, the shares of total resources available for public health were distributed in the following way: Ministry of Social Protection programs, operation expenses, and National Health Institute, 10 percent; PBS, 25 percent; health promotion and preventive individual services included in the contributive and subsidized benefits packages, 17 percent and 35 percent, respectively; and other health services financed by the health promotion and diseases prevention sub-fund for the contributory regime, 14 percent. As of 2004, about one-third of resources were allocated to subnational governments for public health interventions included in the PBS. About 57 percent of total resources were potentially available for individual public health activities provided by insurance, mostly for the contributory regime, given the relative size of this program (Dirección General de Salud Pública, 2004). In summary, pre-reform health spending concentrated on curative care, and levels of financing for public health before 1993 were low and unpredictable. Since 1993, resources available for public health have increased and minimum levels are guaranteed. A large proportion of funds remain tied to individual interventions, however, and resources are fragmented among different agencies and levels of government, which complicates the flow, articulation, and accountability of funding. Trends in the Burden of Disease The bulk of the demographic transition in Colombia occurred during the 1980s. During that decade, large drops in the fertility and mortality

7

As of January 2009.



rates occurred. By 1990, the burden of disease was concentrated in noncommunicable disease. After the mortality drops observed in the 1980s, the pace of the transition slowed. In the period covering the health and decentralization reforms, these trends simply become more pronounced, with the combined share of communicable, maternal, perinatal, and nutritional conditions shrinking further to 17.8 percent of total illnesses. The share of non-communicable diseases increased dramatically—from 39 percent of the total burden to 52.1 percent in 2002. Although a study of avoidable mortality found a leveling out of the rate beginning in 1991 (Gómez, 2005), it is difficult to attribute these patterns to the effects of the insurance or decentralization reforms. Such plateaus are observed worldwide—the kinds of interventions required to reduce infant mortality from a rate such as 18.7 per 1,000 live births require different investments than interventions used when the infant mortality rate was 26 per 1,000 and higher. Trends in Mortality The infant mortality rate (the number of deaths at less than age 1 per 1,000 live births) is a commonly used measure of population health and well-being, and is a gauge of inequalities in access to the public health care system. The infant mortality rate in Colombia decreased from 56.7 per 1,000 in 1975–80 to 18.7 per 1,000 in 2000–05 (Flórez, 2000; Profamilia, 2005). Nonetheless, geographic and economic disparities persist. Not surprisingly, infant mortality in Colombia is generally higher in rural areas, in departments with lower levels of development, and among the poor. This reality can be explained in part by determinants of morbidity, including differences in access to health services, infrastructure, basic services, housing quality, and education. Urban–rural inequalities in infant mortality actually increased between 1995 and 2000, but then decreased during the 2000–05 period. In 2000, infant mortality in rural areas was about 50 percent greater than in urban areas. That difference had decreased to 30 percent by 2005, but the differential in 2005 remained higher than that of 1995. Regional inequalities are also pronounced. In the Pacific region, for example, where the country’s Afro-Colombian population is concentrated,

29



and López

infant mortality is 28 percent higher than it is in urban areas (Flórez and Ruiz, 2006). In the past 10 years, infant mortality rates have fallen faster among the lowest-income quintile than in the richest, changing from 2.5 times higher among the poor to 2.2 times higher. Disparities in infant mortality by health insurance status in 2005 show that mortality rates are highest among uninsured people, slightly lower among enrollees in the subsidized regime, and much lower among those enrolled in the contributory regime (Figure 2.2). This differential has become more pronounced in recent years, given the decrease in infant mortality among all insured people, while infant mortality among the uninsured has increased (from 25.2 to 27.4 per 1,000 between 2000 and 2005). It should be noted that over this same period, the percentage of the population that was uninsured shrank—from 46 to 33 percent of the total population. The differentials in infant mortality across the various health insurance status groups may be partly associated with disparities in access to maternal-infant care according to insurance status. Other factors, such as the impact of the internal armed conflict on children living in affected municipalities, also influence the infant mortality rate (Box 2.1). Figure 2.2 Infant Mortality Rates by Insurance Regime, 2000–05 35 30

TMI (‰)

25 20

Contributory Subsidized Unaffiliated

15 10 5 0

2000 Source: Flórez and Soto (2006).

2005



Box 2.1 Infant Mortality Rate and Violence in Colombia Although the infant mortality rate (IMR) in Colombia has decreased from 32.3 per 1,000 live births in 1986 to 18.7 per 1,000 in 2005 (Demographic and Health Survey data), the country has experienced an even faster decline in fertility levels and improvement in other socioeconomic variables. Urdinola (2004) examined the extent to which the armed conflict in Colombia may have affected infant mortality during the 1990–2000 period. She first tested the notion of negative effects of violence on infant survival in a cross-country fixed-effects model. She found that if Colombia had had the same socioeconomic profile as the other 10 Latin American countries, by the year 2000 it would have had, on average, an IMR of about 5 deaths per 1,000 live births lower than it actually had. Second, Urdinola estimated a fixed-effects model using Colombian municipal data with about 90 percent national coverage to test the effects of violence and other covariates on the IMR. Violence was measured by both homicide rates and armed activities perpetrated by the four guerrilla groups active in Colombia, paramilitary armies, and unorganized criminals. Municipal characteristics include a wide range of variables, including per capita public investment in health and health-related sectors. In the model that used homicides as the proxy for violence, Urdinola found that a reduction in the homicide rate of 1 per 1,000 inhabitants correlated with a decrease in IMR of 3 per 1,000 per year and an additional reduction of 1.5 infant deaths from the lagged effect of violence. What type of violent act mattered most? When measuring violence in terms of armed acts, increased struggle for territory was correlated with an increase in IMR of more than 28 per 1,000 births. An increase in the destruction of physical infrastructure per 10,000 inhabitants coincided with an increase in IMR of 12 per 1,000, the presence of massacres by 3 per 1,000, and increased weapons trafficking by 1 per 1,000. Although the author could not determine causal pathways (e.g., that violence deteriorates the health care system, thus increasing infant mortality; that violence implies an overall lack of State presence and services; or simply that displaced status negatively affects infant survival rates), the relationship between violence and infant survival is clear.

Program Case Studies Vaccination Vaccination programs began in Colombia in 1968, using a combination of campaign and routine strategies administered by a centralized program. Vaccination rates in the 1970s and 1980s were reportedly low (around 20 percent for completed courses of multiple-dose vaccines), in spite of “sufficient human, physical, and economic resources” (Ministerio de Salud, PAHO, 1982). After the reforms of the early 1990s, functions within the vaccination program were split. To realize economies of scale in procurement,

31



and López

vaccines continued to be purchased for the entire country at the national level. The national administration also set norms and policies with respect to the program, and provided technical assistance and limited supervision to sub-national governments. Departments were responsible for supervision, technical assistance, surveillance, and reporting, while municipalities assumed most of the operations of the program, including ensuring that supplies reached providers, monitoring and reporting on vaccination, and conducting campaigns. Insurers were responsible for purchasing syringes and providing vaccinations to insured residents. According to administrative data reported to the Pan American Health Organization, vaccination rates for individual vaccines started at around 16 percent of children under age 1 in 1980 and reached their apex in 1996, with all vaccines in the Expanded Program of Immunization being provided to approximately 95 percent of children under 5 years old (López Casas, 2007). In 1998–99, there was a 15 percentage point drop in vaccination rates, coinciding with the worst economic recession in the country’s history, followed by a recovery from 2000 to 2004. Although spending on PBS was protected by earmarked funding during the recession, levels of vaccination over that period seem to be directly related to the availability of non-earmarked national financing for vaccine purchases (Figure 2.3). Central government funding for vaccine purchases is marginal—between US$25 million and US$35 million per year, or less than one-quarter of one percent of the GDP. In the context of a heavily earmarked total budget at the national level (experts estimate that 85 percent of the national budget is earmarked for salaries and pensions for civil servants), funds for vaccination represented “flexible” spending, vulnerable to cuts as revenue dropped. These cuts illustrate that in spite of the PBS and insurance earmarks, the marginality of the amount represented by vaccine purchases, Colombia’s status as a middle-income country, and the fragmentation in the program’s essential functions resulted in a drop in vaccination rates during the recession. The shortage of the one essential input for the program— vaccines—also resulted in inefficiency in the use of PBS resources at the sub-national level. However, the movement of vaccination rates with vaccine purchases is positive, in that it shows that when inputs are available, the system is able to deliver vaccinations.



Figure 2.3 Vaccination Coverage and Central Government Spending on Immunization, 1998–2003

80

1.000

Expenditure (left scale) Average immunization rate (DPT, Polio, BCG, Hep B, HiB and TV)

0.900

60 0.800 40 0.700

20

Immunization rate (average)

Expenditure (billions of pesos)

100

0.600

0

1998

1999

2000

2001

2002

2003

Source: Authors’ estimates based on Ministry of Social Protection data.

Vaccinating children on time is a major challenge. According to population-based rates derived from a series of demographic and health surveys, complete age-appropriate vaccination for tuberculosis, DPT or equivalent, polio, and measles has declined over time. In 1990, 67.5 percent of children were fully vaccinated with an age-appropriate scheme before age 1; this number was 58.1 percent in 2005.8 Beyond the financing issues associated with the recession, problems in the vaccination program are attributed to the fragmentation of its functions; other observers point to the negative impact of the internal armed conflict on access to poor municipalities as an explanation for low vaccination rates (Gómez, 2005). Others opine that the vaccination program has never been satisfactory and depends on campaigns (Restrepo Trujillo, 2004) to make up for low coverage delivered through routine channels. Several authors, notably Ayala Cerna and Kroeger (2002), have attempted to link the poor performance of the vaccination program 8 Reports only complete vaccination schemes recorded on vaccination cards seen by interviewers; as such, this is a conservative estimate. ORC Macro, 2007. Measure DHS StatCompiler: http:// www.measuredhs.com, June 27, 2007.

33



and López

to the introduction of insurance and managed competition. These studies relied on trends before and after reforms to conclude that the insurance reform had a negative impact on vaccination rates. However, in the only study that uses appropriate methods to establish causality, Giedion et al. (see Chapter 3) used propensity score matching and a quasi-panel of cross-sectional data covering a decade and found that the likelihood of complete vaccination is significantly higher for insured children (6 percent). This trend is more pronounced among rural residents (12 percent). Although vaccination is included in the PBS and thus theoretically available to the entire population, this finding suggests that the greater use of health services resulting from insurance is increasing the likelihood of routine health care visits for children and thus timely vaccination. Beyond insurance status, socioeconomic status (measured by a wealth index) and rural residence affect the equity of vaccination coverage in Colombia. Differentials in vaccination by socioeconomic status worsened after the recession; the wealthiest quintile had vaccination rates 32 percent higher than the poorest quintile in 2000 and these differences have been maintained over time (Flórez and Soto, 2006). Given that vaccination is free and universal, and that geographical access to public care providers is nearly universal in Colombia, these socioeconomic differentials in access may be explained by remaining economic, socio-cultural, and informational barriers to access, including the cost of transportation, opportunity costs, and household knowledge. To respond to these inequities, in 2001 the Government of Colombia implemented a conditional cash transfer program intended to stimulate demand for preventive health care. The program now reaches over 700,000 extremely poor or displaced families. A quasi-experimental impact evaluation, the results of which were published in 2005, found that the program has significantly increased the probability of adequate DPT vaccination for children less than 24 months of age (Attanasio et al., 2005). Malaria The Malaria Eradication Service was established in 1957 as a unit of the former Ministry of Health. It was a centralized vertical program with



its own budget and personnel. Under the new framework created by the health sector reforms, the unit was decentralized. Responsibilities for vector control were delegated to departments, and diagnosis and treatment of malaria to municipalities. In accordance with the Global Malaria Control Strategy and the principles of the Roll Back Malaria Partnership, the Ministry of Social Protection launched the National Malaria Control Program (NMCP) in 1998. The program includes: 1. improved diagnosis and treatment; 2. selective vector control, including use of insecticide-treated nets or mosquito-repellant chemicals; 3. mosquito breeding control and targeted indoor residual spraying; 4. strengthening of public health surveillance, including entomological and vector resistance surveillance; and 5. inter-sectoral and social participation (Korenromp et al., 2005). Currently, Colombia has one of the higher malaria incidences in this region of the Americas, accounting for 10 to 20 percent of cases. The incidence of malaria has been increasing since the 1960s, although there have been larger increments during the past decade, with a peak in 1991 and another in 2002, as depicted in Figure 2.4. Figure 2.4 Annual Malaria Parasite Index, 1960–2002 12 10 8 6 4 2

2 19 64 19 66 19 68 19 70 19 72 19 74 19 76 19 78 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02

19 6

19 6

0

0

Source: Carrasquilla (2006), based on National Institute of Health data (2003).

35



and López

The observed malaria trend in Colombia is likely related to several factors, including climatic changes, increasing resistance to antimalarial medications, resistance of mosquito vectors to insecticides, and internal migration due to the armed conflict (Carrasquilla, 2006). The majority of cases, however, are concentrated in municipalities located in deprived areas, which are not covered by the NMCP because of security concerns (Korenromp et al., 2005). It is also important to note that after the implementation of the NMCP in 1998 there has been an improvement in disease registry, increasing diagnostic coverage by almost 30 percent in areas with high transmission rates (Dirección General de Salud Pública, 2004). Another hypothesis is that the institutional changes brought about by the health sector reforms have affected the implementation of malaria prevention and control measures. Carrasquilla (2006) explored this relationship by compiling secondary data on epidemiological indicators, risk factors, and financial resources for malaria prevention and control in 255 malaria-endemic municipalities in Colombia for the period 1991 to 2000. (The study sought to collect information on cases, hospital discharges, and deaths, and on malaria prevention and control activities, in 319 municipalities. Owing to large gaps in available information, data of varying completeness for each year was obtained from only 255 municipalities.) Carrasquilla conducted semi-structured interviews with relevant health sector officers involved in malaria control regarding operational aspects of the program before and after decentralization. Trends in malaria morbidity were analyzed for three periods: 1990–93 (pre-program decentralization), 1994–98 (transition), and 2000–01 (program decentralization). The study found no statistically significant differences in mean malaria incidence rates among these periods. Because of gaps in information, it was not possible to use a uniform model to explore associations between malaria rates in each period and factors such as climatic variables (including rainfall), decentralization status, insurance coverage, municipal development, and rural population, thus limiting conclusions about the possible determinants of the observed trends. (The study found important gaps in information, such as the absence of data on available resources and expenditures specific to the malaria program, and



a large variability in reported cases in some areas. For example, for the 1994–98 period only 67 municipalities had information about 1997 health expenditures.) Carrasquilla’s findings with respect to the institutional aspects of the malaria program circa 2001 indicate that sub-national governments were allocating resources to, planning, and executing activities in accordance with their responsibilities for malaria control. However, the participation of private institutions in diagnosis and treatment activities is small in relation to the activity of the public sector. Weaknesses that could jeopardize the effectiveness of the malaria program identified by the survey include frequent rotation of personnel and lack of training of officers appointed for malaria control. Tuberculosis As in the cases of vaccination and malaria, prior to 1993 the tuberculosis control program was a national vertical program that comprised planning, administration, technical assistance, financing, and provision of care through public hospitals. Tuberculosis services were offered free throughout the country. After the insurance and decentralization reforms, the Ministry of Health became responsible for policies, norms, and procurement of first-line medications and the anti-tuberculosis vaccine Bacillus Calmette-Guérin (BCG). Departmental health directorates provided technical assistance, monitoring, supervision, and distribution of medicines and vaccines to municipalities, while municipalities were responsible for providing PBS services, which included monitoring tuberculosis control activities, distribution of medicines and vaccines to providers, carrying out home visits, and providing treatment to uninsured patients. Insurers provided vaccines to their populations and, after Law 715, referred tuberculosis patients to the public sector for treatment. A review of tuberculosis incidence published in 2004 reported an incidence rate for all forms of tuberculosis that declined from 34 cases per 100,000 population in 1992 to approximately 26 cases per 100,000 in 2002 (Chaparro et al., 2004). A rate calculated by the authors based on the number of detected cases from the National Health Institute’s

37



and López

epidemiological surveillance system9 and denominators constructed from the 2005 census indicates a national average that has oscillated around 20 per 100,000 from 2003 to 2006. Although there are differences in the incidence rates reported by the ministry (26 per 100,000 in 2005) and the National Health Institute (24 per 100,000 in 2005), the number of notified cases in each series has remained relatively steady over this period, ranging from 8,308 cases in 2003 to a high of 9,009 cases in 2004, declining again to 8,300 cases in 2006. The World Health Organization reports a very different rate of 45 per 100,000, based on officially notified cases adjusted for estimates of under-reporting (WHO, 2007). While some authors give importance to the slight increase in cases observed in 2002 (Ayala Cerna and Kroeger, 2002), the tuberculosis rate picture is unclear. It may have worsened or stayed more or less the same over the decade. Stable tuberculosis mortality rates and declining hospital discharge rates observed in the late 1990s have been interpreted by some as evidence of stagnation (Segura, Rey, and Arbeláez, 2004). Factors hypothesized to explain tuberculosis incidence trends include those related to fragmentation of care provision, which is said to have led to late diagnosis, more frequent hospitalization, and higher mortality (Arbeláez, 2006). Although no studies have rigorously analyzed the impact of insurance status on case detection and treatment, insurance status seems to affect adherence to treatment: a 1999 study in Bogotá of 726 cases found that adherence was higher among the insured in the contributory regime and lowest amongst the uninsured (Arbeláez, 2006). Treatment errors by health providers also play a role; in small-scale municipal studies, such errors have been observed, leading to modest levels of drug-resistant tuberculosis (Moreira et al., 2004; Laserson et al., 2000). Co-infection with human immunodeficiency virus (HIV) is also hypothesized to affect tuberculosis status, although only 5 to 10 percent of diagnosed tuberculosis cases present with HIV co-infection (Chaparro et al., 2004; García et al., 2004). Application of directly observed treatment short-course (DOTS) is considered low; in 2005, the WHO

9

www.ins.gov.co, accessed June 2007.



estimated that 50 percent of cases benefited (Dirección General de Salud Pública, 2000). In contrast to these findings, a regional study published in 2003 indicated that Colombian incidence rates were a result of better implementation of DOTS (Sobero and Peabody, 2006). BCG coverage is high, with 97 percent of children under 4 years old having received the vaccine, consistent with high rates of professionally attended births (Profamilia and Macro International, 2006).

Discussion Overall, public health status has improved in Colombia, although inequities remain. The evidence indicates that spending on public health has increased substantially and that insurance increases access to some key interventions (see Chapter 3). Given that, a more rapid transition to full insurance is an important vehicle for improving public health. The ongoing armed conflict and the recession of the late 1990s have played important roles in explaining results observed in public health programs. Just as these factors explain the slow extension of the subsidized health insurance regime (see Chapter 6), the fall in vaccination rates appears directly linked to the budgetary effects of recession: less vaccine was purchased and fewer children were vaccinated. Urdinola’s study (2004) of the impact of violence in certain municipalities as a significant determinant of the pace of infant mortality rate decline also illustrates the sometimes limited scope of health sector interventions. The inability of the vaccination and malaria programs to work in certain highly vulnerable municipalities for security reasons also limits the impact of the programs. More can be done to isolate the importance of these multiple forces affecting public health outcomes, leading to more nuanced policy options, and measures could be taken to ensure that essential public health inputs are protected during periods of economic downturn. Among the most critical challenges facing public health is the fragmentation of health care functions among levels of government. This fragmentation was created by decentralization, combined with a lack of articulation among the different participants in the insurance system. A vaccination program that puts one organization in charge of procuring vaccines, several others in charge of procuring syringes,

39



and López

and yet others in charge of contracting vaccinators is likely to show poor performance because the inputs necessary to vaccinate on time are simply unavailable. Although recent government efforts have sought to better align incentives, much remains to be done to adjust those incentives to improve the impact of the program. For example, the Ministry of Social Protection has consolidated the procurement of vaccines and syringes with one entity and now regularly supervises insurers and municipalities to ensure timely vaccination. The lack of unified and effective stewardship and accountability for public health outcomes remains problematic. Slow and partial responses to outbreaks of diseases such as dengue fever are an example: a 2004 report of an outbreak of dengue found that only 1 of every 9 suspected dengue cases presenting at emergency rooms was reported (Loevinsohn and Harding, 2005). As a result, national funding and technical assistance to deal with the outbreak arrived late and avoidable deaths occurred. Poor surveillance was attributed to a complex, facility-based reporting procedure, which has now been complemented by the implementation of a sentinel surveillance system. Health workers specializing in the control of communicable diseases have also reported reassignment to other functions by municipal or departmental health authorities, indicating limited capacity to understand the issues at stake, particularly in the poorer municipalities. Since much of the communicable disease occurring in Colombia is concentrated in a core number of poor municipalities, efforts and financing could be targeted more effectively, while still operating within the framework of reform and decentralization. While popular throughout Latin America, the conceptual model that separates individual and collective health interventions should remain conceptual. Its enshrinement in legislation, financing, and the content of benefits packages has unnecessarily complicated careseeking and interrupted the continuum of care. Although its intention was the opposite, and it was later revoked in Law 1122 of 2007, Law 715 aggravated this situation by removing key prevention and promotion interventions from insurance packages and making municipalities exclusively responsible for their provision. Future efforts should seek to establish benefits packages and associated financing arrangements that facilitate care-seeking and adherence to treatments, no matter



what the economic rationale for investment or the insurance status of the affected individual. Colombia’s experience confirms that governance conditions are important to the effectiveness of health care programs. Municipalities were allowed to contract out the contents of the PBS, yet many opted to execute directly or carry out direct contracts that were vulnerable to misuse and generated disappointing results. Open and competitive contracting for key public health services has shown promise elsewhere in the world (Loevinsohn and Harding, 2005), yet this potentially innovative policy opportunity was lost in many Colombian municipalities and resulted in misuse of funds and limited impact of services. The new requirement to use a portion of public health funds to contract with public hospitals worsens the situation. Poor-quality data and research mean limited policy effectiveness. From the example of the outdated census to the uncertainties around the incidence rate of tuberculosis, it is difficult to design policies and ensure their intended results in the absence of at least minimal data. Further, research methods must be strengthened. Many studies reviewed for this chapter lack power, or use inappropriate methods to establish causal links between reforms and outcomes observed, or both. Finally, official data on public health are dispersed and inconsistent. Colombian authorities should do more to ensure the consistency and accuracy of public health data collected and used by institutions in Colombia and reported to international agencies such as the World Health Organization. Tuberculosis incidence rates reported by the WHO are double what any source in Colombia reports. These inconsistencies muddy the policy waters and can lead to spurious conclusions.

41



and López

References Arbeláez, M.P. 2006. La reforma del sector salud y el control de la tuberculosis en Colombia. In Decentralization and Management of Communicable Disease Control in Latin America, eds. Z.E. Yadón, R.E. Gürtler, F. Tobar, et al. Buenos Aires: Pan American Health Organization. Attanasio, O., L.C. Gómez, P. Heredia, et al. 2005. The Short-Term Impact of a Conditional Cash Subsidy on Child Health and Nutrition in Colombia. London: Institute for Fiscal Studies. Ayala Cerna, C., and A. Kroeger. 2002. La reforma del sector salud en Colombia y sus efectos en los programas de control de tuberculosis e inmunización. Cadernos de Saúde Pública 18(6): 1771–81. Carrasquilla, G. 2006. Descentralización, reforma sectorial y control de la malaria. In Decentralization and Management of Communicable Disease Control in Latin America, eds. Z.E. Yadón, R.E. Gürtler, F. Tobar, et al. Buenos Aires: Pan American Health Organization. Chaparro, P.E., I. García, M.I. Guerrero, et al. 2004. Situación de la tuberculosis en Colombia, 2002. Biomédica 24(Sup. 1): 102–14. Dirección General de Salud Pública. 2000. Situación actual de la tuberculosis en Colombia. Bogotá: Ministerio de Salud. ———. 2004. La salud pública en Colombia: análisis y propuestas. Mimeo. Bogotá: Ministerio de la Protección Social. Flórez, C.E. 2000. Las transformaciones sociodemográficas en Colombia durante el siglo XX. Bogotá: Banco de la República. Flórez, C.E., and M. Ruiz. 2006. Análisis de situación para la formulación del programa de cooperación del UNFPA con el país para el período 2008–2012. Report for the United Nations Fund for Population Activities. Flórez, C.E., and V. Soto. 2006. Inequidades en salud en Colombia: 15 años de avances. Bogotá: Fundación Corona, Departamento Nacional de Planeación, Programa de Naciones Unidas para el Desarrollo. García, I., A. Merchán, P.E. Chaparro, et al. 2004. Panorama de la coinfección tuberculosis/VIH en Bogotá, 2001. Biomédica 24(Sup. 1): 132–37.



Gómez, R.D. 2005. La mortalidad evitable como indicador de desempeño de la política sanitaria: Colombia 1985–2001. Doctoral dissertation, Universidad de Alicante, Alicante, Spain. Grupo de Gestión Integral en Salud. 2003–2005. Informe final de la consolidación y análisis de los recursos a cargo del 4,01% de la UPC de Régimen Subsidiado Vigencias. Bogotá: Ministerio de la Protección Social. ———. 2005. Resultado de la Gestión del Programa de Enfermedades de Transmisión por Vectores. Bogotá: Ministerio de la Protección Social. Herrera, V., and A. Cortez. 2000. Análisis de la descentralización de la política social y la municipalización del sistema general de seguridad social en salud en municipios de categoría 4, 5 y 6. Bogotá: ESAP-CINAP. Homedes, N., and A. Ugalde. 2005. Las reformas de salud neoliberales en América Latina: una visión crítica a través de dos estudios de caso. Revista Panamericana de Salud Pública 17(3): 210–20. Jaramillo, I. 1999. El futuro de la salud en Colombia: cinco años de la puesta en marcha de la ley 100, 4th ed. Bogotá: FESCOL, FES, FRB, Fundación Corona. Korenromp, E., J. Miller, B. Nahlen, et al. 2005. World Malaria Report 2005. Geneva: World Health Organization and UNICEF. Laserson, K., L. Osorio, J. Sheppard, et al. 2000. Clinical and Programmatic Mismanagement Rather Than Community Outbreak as the Cause of Chronic, Drug-Resistant Tuberculosis in Buenaventura, Colombia, 1998. The International Journal of Tuberculosis and Lung Disease 4(7): 673–83. Loevinsohn, B., and A. Harding. 2005. Buying Results? Contracting for Health Service Delivery in Developing Countries. The Lancet 366(9486): 676–81. López Casas, J.G. 2007. La inmunoprevención en Colombia 1980–2006. Unpublished mimeo. Ministerio de Salud; Pan American Health Organization. 1982. Evaluación del Programa Ampliado de Inmunizaciones. Bogotá: PAHO. Molina, C.G., U. Giedion, M.C. Rueda, et al. (FEDESARROLLO). 1994. El gasto público en salud y distribución de subsidios en Colombia. Informe final. Bogotá: Departamento Nacional de Planeación,

43



and López

Misión de Apoyo a la Descentralización, Focalización de los Servicios Seccionales. Moreira, C.A., H.L. Hernández, N.L. Arias, et al. 2004. Initial Drug Resistance as a Threat for Tuberculosis Control: The Case of Buenaventura, Colombia. Biomédica 24(Sup. 1): 73–79. Profamilia. 2005. Salud sexual y reproductiva en Colombia, Encuesta Nacional de Demografía y Salud 2005. Resultados StatCompiler. Profamilia and Macro International. 2006. Encuesta Nacional de Demografía y Salud, Colombia 2005. Calverton, MD: Profamilia and Macro International. Restrepo Trujillo, M. 2004. A New Reform of the National Health System. Biomédica 24(4): 341–44. Richards, T. 1997. Colombia Struggles with Health Reform. British Medical Journal 315(7107): 501–04. Sánchez, L.H., and F.J. Yepes. 1999. La descentralización de la salud en Colombia: estudio de casos y controles. Informe técnico. Bogotá: Asociación Colombiana de la Salud. Sánchez, L.H., F.J. Yepes, and B. Cantor. 1998. La descentralización de la salud: el caso de tres municipios colombianos. Informe técnico. Bogotá: Asociación Colombiana de la Salud. Segura, A.M., J.J. Rey, and M.P. Arbeláez. 2004. Tendencia de la mortalidad y los egresos hospitalarios por tuberculosis, antes y durante la implementación de la reforma del sector salud, Colombia, 1985–1999. Biomédica 24(Sup. 1): 115–23. Sobero, R.A., and J.W. Peabody. 2006. Tuberculosis Control in Bolivia, Chile, Colombia and Peru: Why Does Incidence Vary So Much Between Neighbors? The International Journal of Tuberculosis and Lung Disease 10(11): 1292–95. Tono, T.M., L. Velásquez de Charry, J. Sáenz, et al. 2002. El impacto de la reforma sobre la salud pública: el caso de la salud sexual y reproductiva. Bogotá: Fundación Corona, Fundación Ford, Engender Health. Unión Temporal CCRP-ASSALUD-BDO. 2004. Diseño y aplicación de una encuesta para la evaluación de las acciones de prevención del POS-C/POS-S y del logro de las metas del PAB departamental y distrital. Bogotá: Unión Temporal CCRP-ASSALUD-BDO.



Urdinola, P. 2004. Could Political Violence Affect Infant Mortality? The Colombian Case. Unpublished manuscript. University of California, Berkeley. Vargas, J.E., and A. Sarmiento. 1997. Descentralización de los servicios de educación y salud en Colombia. Bogotá: Casa Editorial El Tiempo. Vivas, J., E. Tarazona, C. Caballero, and N. Marrero. 1988. El Sistema Nacional de Salud. Administración, presupuestación, gasto y financiamiento, 1st ed. Bogotá: FEDESARROLLO, Pan American Health Organization. World Health Organization. 2007. WHO 2005 Tuberculosis Epidemiological Profile—Colombia. Geneva: WHO. ———. WHO Statistical Information System. Available at http://www. who.int/whosis/en/index.html. Accessed June 2007.

45




The Impact of Subsidized Health Insurance on Health Status and on Access to and Use of Health Services Ursula Giedion, Beatriz Yadira Díaz, Eduardo Andrés Alfonso, and William D. Savedoff

I

n the early 1990s, Colombia introduced a universal health insurance scheme with two forms of affiliation. The contributory regime covers formal sector workers earning at least one minimum salary (about US$223) per month,1 and informal and independent workers earning at least two minimum salaries per month; the subsidized regime covers those considered poor according to a proxy means test, the Sistema de Identificación de Beneficiarios (SISBEN; Beneficiary Identification System). Individuals who qualify for the contributory regime are charged a 12 percent payroll tax for a comprehensive insurance plan valued at about US$207.2 Payroll tax contributions are pooled by a public

1 Minimum salary for 2007, defined by Decree 4580/2006 of the Ministry of Social Protection, was $Col 433,700, equivalent to US$223. Exchange rate at June 15, 2007: $Col 1,945/US$1. 2 Premiums for 2007 were established by Agreement 35/2006 of the National Social Security Council at $Col 404,215.20 (contributory regime) and $Col 227,577.60 (subsidized regime). Dollar values, using the exchange rate of June 2007 ($Col 1,945/US$1) were US$207 (contributory regime) and US$117 (subsidized regime).


chapter 3

and Savedoff

fund, the Fondo de Solidaridad y Garantía (FOSYGA; Solidarity and Guarantee Fund), which channels resources from individuals whose contributions are greater than the value of premiums for themselves and their families to those whose contributions are less. For those who cannot afford to purchase insurance, the government uses national revenues, local revenues, and a portion of the payroll tax (1 percent of payroll) from the contributory regime to purchase insurance coverage for the poor under the subsidized regime. The benefits package is more limited in the subsidized regime (costing about US$117) but legislation calls for it to become similar to the contributory regime, depending on the mobilization of additional resources. By 2007, most basic care and most high-cost interventions related to catastrophic illnesses such as cancer and acquired immunodeficiency syndrome (AIDS) were covered under the subsidized regime. Most hospital care is therefore not yet covered; for these services, rules of access do not differ for insured and uninsured. Under either regime, the patient chooses a health insurance company, which may be public, private, or mixed and which may be run for profit or not for profit. The insurance company, in turn, covers a portion of health care costs by establishing contracts with public and private providers or through its own health care providers. Insurance companies are paid a risk-adjusted per capita amount. As a result of these reforms, insurance coverage increased from 24 percent of the population in 1993 to 62 percent in 2003. The increase was largest among the lowest-income quintile, rising from 6 percent before the reforms to 47 percent a decade later. The current government intends to achieve universal coverage during its term by mobilizing new financial resources. Despite these gains, criticism of the reforms is common. Several opposition groups have called for massive changes to the system; the reforms have been prominently debated in the past two presidential elections. Although growing empirical evidence exists on the benefits of the subsidized health insurance scheme—specifically, access to and utilization of care (see, for example, Panopoulou, 2001), financial protection for households against out-of-pocket costs (for example, Flórez, Giedion, and Pardo, 2007), and better targeting of public-sector resources (Acosta et al., 2007)—many argue that the health system was


48 Giedion, Díaz, Alfonso,

better before the reforms and claim that the new system has worsened health conditions. Further, some observers consider that the large fiscal effort involved in financing the subsidized health insurance scheme (about 1 percent of gross domestic product in 2003; Barón, 2007) may not be worthwhile and may have had a negative impact on employment (Gaviria, Medina, and Mejía, 2006). This chapter provides evidence to inform and enrich such debate and to call attention to the risks involved when supporting massive changes before considering how they could alter the positive results already achieved. Finally, Colombia has been one of the first countries in the developing world to introduce a social insurance scheme providing universal coverage and equal financial access to a basic benefits package for all (Panopoulou, 2001). The expansion of insurance coverage among the poor has been on the agenda of many countries and international organizations as a means of improving access to care and financial protection for those most vulnerable to the consequences of illness. Therefore, by reaching almost two-thirds of its population with insurance coverage, Colombia’s case provides a unique opportunity to gather evidence on one of the most hotly debated issues in the health sector. In this context, evidence of the impact of the Colombian health reforms is urgently needed, not only to inform policymaking in Colombia but also to provide lessons for other countries considering similar reforms. This study uses existing data and impact evaluation methods to measure the effects of the Colombian subsidized regime on the levels and distribution of insurance coverage, health service access and utilization, and health status. It confers robustness to its results by combining and comparing the results from several different semiparametric impact evaluation methods.

Background and Context Why Care about Insurance? Health insurance reduces the direct costs of access to and utilization of health care services by individuals and families. It therefore reduces the financial risk of illness and improves access to health services. This

49


Impact of Subsidized Health Insurance on Health Status, Access, and Use

and Savedoff

study addresses the hypothesis that the subsidized health insurance regime introduced in Colombia in 1993 has improved the health status of the insured population by making health care more affordable. Several qualifications are in order.3 First, health insurance affects only the affordability of health care; it does not necessarily alter the other factors that affect access.4 Second, health insurance affects health indirectly through its impact on health care utilization. Third, the effect of health insurance may vary across the population. In particular, in areas with effective social safety nets, lack of insurance may not be a significant barrier to receiving care and, consequently, the marginal impact of introducing insurance coverage may be small compared with the impact in areas where individuals have fewer options (Buchmueller et al., 2005). Fourth, people who have health insurance may differ systematically, in some consistent way, from those without insurance, making analysis more difficult. Finally, health status is itself a complex concept and findings may vary depending on the particular variables chosen to measure it. Eligibility for and Affiliation with the Subsidized Regime Participation in the subsidized regime is a two-step process: according to the existing legal framework, the vulnerable population is first identified as being eligible and then gradually affiliated with the subsidized regime based on several predefined prioritization criteria. To model “participation”—a key issue when using quasi-experimental methods such as propensity score matching or matched difference-in-differences; it is used in this impact evaluation of the subsidized regime—it is necessary to understand what determines how and why an individual becomes eligible for subsidized health insurance, and what determines whether an eligible person is affiliated with the subsidized regime. We will briefly discuss these issues below.

For an excellent review of these issues see Buchmueller et al. (2005) and Levy and Meltzer (2001). 4 Penchansky and Thomas (1981) identify five dimensions of access: availability, accessibility, accommodation, affordability, and acceptability, as described in McLaughlin and Wyszewianski (2002). 3



Eligibility rules for enrollment in the subsidized regime are complex. This complexity affects the analytical strategy employed in the study. First priority is given to special populations such as orphans and the elderly, irrespective of proxy means test scores. Priority is then given to the poor with low test scores who are either pregnant, under the age of 5, displaced by violence, or disabled. The remaining population is ranked according to scores obtained under the SISBEN. After this ordered list is published, selected individuals can subscribe to one of the competing insurance entities. If an individual does not sign up, he or she loses the opportunity to enroll in the insurance scheme and must wait for the next round of affiliations. Affiliation of those eligible occurs gradually as additional funds become available nationally and locally. Those eligible but unaffiliated can use public hospitals at highly subsidized prices but are not granted the explicit and legally guaranteed benefits package of those who are insured. The system’s implementation introduced further complications. First, the proxy means test and affiliation were introduced unevenly across the country, depending on the availability of additional local funds and municipal administrative capacity. Second, the distinction between those with and those without insurance is somewhat blurred, given that the latter are granted partial fee waivers in public hospitals (see Panopoulou, 2001). Third, some evidence indicates that SISBEN scores and affiliation are manipulated by local authorities, leading to the inclusion of non-poor populations. (Despite the limited coverage and some leakage of subsidies to wealthier people, the subsidized regime is still the best-targeted social program in Colombia and the health sector has made the most progress with targeting in the past two decades. See Lasso, López, and Núñez, 2004.) In summary, legislation guiding participation in the subsidized regime, data from previous surveys, and analyses carried out by several researchers indicate that participation in the subsidized regime is not random and depends on many variables other than poverty scores (Panopoulou, 2001; Trujillo and Portillo, 2005). Therefore, simple comparisons of differences in outcome between affiliates and non-affiliates would most certainly yield biased estimates of the impact of subsidized health insurance in Colombia.

51



and Savedoff

Previous Evidence of the Impact of Subsidized Health Insurance in Colombia A number of studies have analyzed the impact of the subsidized regime on utilization of health care services, financial protection, and health status. Studies using data corresponding to the first stage of implementation of the subsidized regime find evidence of the positive impact of health insurance for outpatient care but not for hospital care (Panopoulou, 2001; Trujillo and Portillo, 2005).5 The latter result can be explained by the limited coverage of hospital services under the subsidized regime, as previously indicated. Both reports find stronger evidence of a positive impact in urban areas than in rural areas. Relying on more recent data, Gaviria et al. (2006) found a positive and substantial impact of the subsidized regime on the use of preventive medical care and outpatient visits, and a negative impact on hospitalization rates at the national level. According to these authors, the former result may be explained by the fact that uninsured patients have higher emergency and, consequently, hospital utilization rates. Bitrán et al. (2004) use descriptive statistics to show that poor insured people under the subsidized regime benefit from lower rates of unsatisfied demand and fewer financial barriers when accessing services, make more outpatient visits, have lower out-of-pocket health care spending, and have a lower incidence of catastrophic health expenditure than do poor people lacking insurance. Note, however, that the former results are based on the comparison of simple means and may be biased, owing to potential differences between affiliates and non-affiliates. Only a few studies have sought evidence of the impact of the subsidized regime on health status. Gaviria and his colleagues use self-reported health status and birth weight as health status outcome measures (Gaviria et al., 2006; Gaviria and Palau, 2006). They found that insurance has a positive impact on health status perception using an instrumental variable approach, but given the method these authors chose, this result requires a questionable leap of faith with respect to the independence of health status perception (outcome) from social 5 Both studies use Living Standards Measurement Study data from 1997, which corresponds to the first years of implementation of the subsidized regime.



and political context at the municipal level, as measured by the number or share of years lived in the same municipality (the “instrument”).6 These authors also find that birth weight increases slightly for insured patients but only for those belonging to the very poorest strata of the population. It is, however, impossible to tell whether insured babies are healthier just because they weigh 50 grams more than uninsured babies. Only when weight falls below a certain threshold is a child’s health at risk. This is why many authors use low birth weight, or extremely low birth weight, as a proxy when evaluating the impact of health insurance on infants. Some of the difficulties in the earlier literature are caused by bias. Researchers applied different methods, ranging from descriptive analysis to instrumental variables and semi-parametric approaches, to address bias. In all cases, researchers had to struggle with questions related to the difficulty of interpreting causality between health insurance and selected outcome variables. The present study complements the existing evidence by: testing the robustness of results through the implementation of several impact evaluation methods; taking advantage of a quasi-panel data set; and combining in one study the analysis of an array of access, utilization, and health status variables not only at the national level but also by poverty level and by area. (Note that none of the previous studies used repeated cross-sectional data instead of cross-sectional data to correct for some of the potential selection problems related to differences between affiliates and non-affiliates in unobserved characteristics.)

Methods When experimental data are unavailable, the choice of analytical approach depends on the specific circumstances and often requires testing several methods (see Blundell and Dias, 2000). To control for selection bias due to differences between affiliates and non-affiliates, and to test the robustness of the results, four different methods were implemented, including a regression discontinuity approach (RDA), propensity score 6 See the complete report (Giedion and Díaz, 2007) for more detail on the instrumental variables approach.

53



and Savedoff

matching (PSM), and matched double difference (MDD). The RDA was discarded because the data showed that one of this method’s central assumptions—the randomness of affiliation based on the proxy means test eligibility score—did not hold in the context of the Colombian subsidized regime (for further details see Giedion and Díaz, 2007). When good panel or repeated cross-sectional data are available, MDD is superior to PSM because it controls not only for differences between affiliates and non-affiliates in observable characteristics (for example, education, income, and housing characteristics) but also for time-invariant differences in unobservable differences (Blundell and Dias, 2000). There was, however, a tradeoff between precision of the estimate and control for selection bias: MDD is inferior to PSM in terms of the richness of outcome variables found in the available data sets. The repeated crosssectional data set required for MDD contained a substantially poorer set of access-, utilization-, and health status–related variables than the cross-sectional data from the Demographic and Health Survey (DHS) from 2005 needed to implement PSM. Results from both methods are presented below.7 Data Description and Sample This study uses a combination of repeated cross-sectional DHS data from 1995, 2000, and 2005; 1993 census data; and municipal administrative data. No source other than the DHS provides adequate and comparable pre- and post-intervention data on individual health status. Administrative data provided contextual variables (such as health services supply, local management capacity, and financial resources) to analyze the determinants of affiliation with the subsidized regime. Census data from 1993 provided additional information on conditions prior to the reforms. For reasons of confidentiality, it was not possible to obtain individual census data, so block data (each block representing approximately 20 households) had to be used instead. Further details on the data, variables, and matching processes can be found in Giedion and Díaz (2007).

7

Results from other methods are available on request.



The sample was drawn from the different rounds of the DHS (1995, 2000, and 2005) and included all individuals affiliated with the subsidized regime as well as those who were uninsured—that is, lacking affiliation with either the subsidized or the contributory regime. Individuals affiliated with the contributory regime were excluded. This restriction excludes the majority of middle- and upper-income individuals from the sample.

Results Impact of the Subsidized Regime on Health Insurance Coverage The increase in health insurance coverage among Colombians is the one successful outcome on which most observers—supporters and opponents of the reform alike—generally agree. It is also an outstanding result at the international level because very few low- and middle-income countries have expanded health insurance coverage so rapidly and to such high levels in such a short time. (Similar coverage levels are being attained in Thailand and the Philippines; Costa Rica and Chile have achieved universal coverage but over a longer time frame.) Overall, health insurance coverage in Colombia has increased from less than a quarter of the population prior to the reforms (1993) to almost two-thirds of the population a decade later (Escobar, 2005). More recent official administrative information indicates that by 2006, 82.72 percent of the population was covered by health insurance either in the subsidized regime (54 percent) or the contributory regime (46 percent) (Ministerio de la Protección Social, 2006). The growth of insurance coverage was most notable among the poorest quintile, where the insured portion of the population increased almost eight-fold (Escobar, 2005). Data from 2005 (Figure 3.1) indicate that the subsidized regime is well targeted to the poor, since its coverage increases with poverty, whereas coverage by the contributory regime increases with wealth (see Chapter 6 for further details on targeting). Coverage is similar for both genders and is somewhat higher among teenagers and those over 50 years old. Differences in coverage

55



and Savedoff

Figure 3.1 Health Insurance Status by Poverty Level (2005) Subsidized regime

Contributory regime

Not affiliated

Doesn’t know

110% 100%

1%

1%

1%

1% 10%

35

90%

21%

30 80% 49%

Affiliation TMI (%o)status

1%

43%

34%

25 70% 60% 20 50%

15

8%

3%

25%

58%

84%

Contributory Subsidized 100% Unaffiliated

40%

10

30%

5

20%

48%

48%

40%

0

10%

2000

19% 2005 5%

0%

1

2 3 4 5 Poverty quintiles as measured by SISBEN (proxy means test)

6

Source: Giedion and Díaz (2007) based on Demographic and Health Survey 2005 data. SISBEN = Sistema de Identificación de Beneficiarios.

depending on municipality are quite substantial: in about one-fifth of Colombia’s municipalities, insurance coverage is still below 20 percent, while two-fifths of municipalities have achieved over 80 percent coverage. This variation is primarily a consequence of inequities in the local resources that are applied to the subsidized regime. Impact of Subsidized Health Insurance on Access to and Utilization of Health Services, and on Health Status Table 3.1 describes the 13 access and five health status variables used in this study, with information on simple unconditional mean differences between individuals affiliated and not affiliated with the subsidized regime and who are at poverty level 3 or below. The data sets contain substantial information on access to services but very little on health



status. Furthermore, the health variables focus exclusively on maternal, newborn, and young children’s health. This complicates the analysis because many health services for small children and childbearing women are free for everyone regardless of insurance status and, therefore, the financial barriers addressed by having insurance coverage are likely to be less important. Moreover, it is not possible to extrapolate from these variables the impact of insurance coverage on other population health conditions, particularly those that can be directly improved by most insured health care services. Comparison of Unconditional Means Comparison of unconditional means of affiliated and unaffiliated individuals belonging to the lowest strata of the population (SISBEN level 3 and below) indicates that those with coverage are less likely to go without care when they need it (26 percent compared with 46 percent for people without coverage; Table 3.1). In addition, only 24 percent of the insured report that their access problems are related to financial barriers, compared to 57 percent of those who are unaffiliated. Instead, affiliated patients more often report difficulties due to limited supply (30 percent, compared with 13 percent for non-affiliates). They use health services more often (68 percent versus 46 percent), and insured small children with coughing or diarrhea are brought more often to a health facility. Differences related to access to prenatal, birthing, and postpartum services are less pronounced but also show significantly better access for pregnant women with subsidized insurance: they receive 4 percent more prenatal visits, take their babies to health facilities 3 percent more often, and are assisted by professionals (4 percent) or doctors (5 percent) more often than are women without subsidized insurance. The impact of subsidized insurance on health status, based on the simple comparison of means, provides mixed evidence: the difference in survival of small children is statistically insignificant. According to information provided on birth certificates, affiliates have a lower incidence of extremely low birth weight (0.3 percent versus 1.4 percent for unaffiliated babies) but a higher incidence of low birth weight

57



13.2%

Not receiving care when needed because of supply problems

30.4%

26.1%

Mean value, affiliates

+130%

−43%

Difference (%)

***

***

Statistical significance

Household members

Household members

Universea


If you did not receive any care when requiring attention, to what circumstances was this due? 1 = Due to supply problems, including services too far away, services of low quality, didn’t attend, didn’t resolve problem last time, too much paperwork 0 = Did not receive care for other reasons

What did you do when having a health problem you considered severe enough to require attention? 0 = Went to health facility, to physician, alternative therapy, nurse 1 = Did not receive any medical care

Variable construction


45.7%

Mean value, non-affiliates

Not receiving medical care when needed (excluding health problems considered too minor to require attention)

Access variables

Variable description

Table 3.1 Access to and Utilization of Health Services, and Health Status (2005)

58 Giedion, Díaz, Alfonso, and Savedoff

46.2%

37.4%

35.7%

Had outpatient visits in past 12 months

Child immunization complete for age

Child taken to health care facility when coughing 44.8%

41.8%

68.2%

23.8%


+26%

+12%

+48%

−58%

Difference (%)

***

***

***

***


Children under 5

Children under 5

Household members

Household members

Universea


Was child taken to health care facility when child had fever or cough? 0 = No 1 = Yes

Is immunization complete for age? 1 = Yes 0 = No

Has any household member used health services in the past 12 months? 0 = No 1 = Yes

If you did not receive care when having a problem considered severe enough to require attention, was this due to lack of money? 0 = No, did not receive care for other reasons 1 = Yes



56.9%


Not receiving care when needed because of financial barriers


Table 3.1 Access to and Utilization of Health Services, and Health Status (2005) (continued) Impact of Subsidized Health Insurance on Health Status, Access, and Use

59

5,190

83.2%

81.3%

Number of prenatal visits

Birth in health facility

Birth attended by professional

84.7%

85.8%

5,393

35.5%


+4%

+3%

+4%

+21%

Difference (%)

***

***

***

**


Women

Women

Women

Children under 5

Universea


Was birth attended by professional? 0 = No = trained birth attendant, relative, friend, other person, other response (uncoded), no one 1 = Yes = doctor, nurse, midwife, auxiliary midwife, health professional

Child delivered in formal setting? 0 = No = home delivery, other 1 = Yes = hospital, clinic, government health center, post– insurance-financed center, private doctor

Number of prenatal visits when pregnant

Was child taken to health care facility when child had diarrhea? 0 = No 1 = Yes



29.4%


Child taken to health care facility when having diarrhea


Table 3.1 Access to and Utilization of Health Services, and Health Status (2005) (continued)

60 Giedion, Díaz, Alfonso, and Savedoff

47.0%

42.3%

Post-natal visit after delivery

Access to medical service when complications present (n = 453 in DHS 2005)

2.9

Health status perception score

2.8

97.2%

48.8%

52.1%

80.0%


−3%

—

+15%

+11%

+5%

Difference (%)

***

***

***

***


Household members

Children under 5

Household members

Women

Women

Universea


How do you perceive your health status? 1 = Not good 2 = Normal 3 = Good 4 = Very good 5 = Excellent

1 = Yes, child is alive 0 = No, child has died

Received medical attention because of complications 0 = No 1 = Yes

0 = No 1 = Yes = doctor, nurse, auxiliary nurse

0 = No 1 = Yes



97.7%

Survival of children younger than 5 years

Health status variables

76.5%


Birth attended by doctor


Table 3.1 Access to and Utilization of Health Services, and Health Status (2005) (continued)


61

31.5%

0.3%

10.9%


+5%

−77%

+43%

Difference (%)

*

**

**


0 = No 1 = Yes = excessive bleeding, loss of consciousness, fever, breast infection, pain when urinating, postpartum depression

0 = No = birth weight > 1500 g 1 = Yes = birth weight ≤ 1500 g

0 = No = birth weight > 2500 g 1 = Yes = birth weight ≤ 2500 g


Women

Children under 5

Children under 5

Universea


Source: Authors, based on Demographic and Health Survey 2005 data. *** = p

From Few to Many

From Few to Many

Many Miles to Go

Many Lives, Many Masters

From Bulk to Nano: The Many Sides of Magnetism

From Bulk to Nano: The Many Sides of Magnetism

The Many Faces of Philosophy: Reflections from Plato to Arendt

So Many Ways to Begin

So Many Ways to Begin

Introduction to many-body physics

So Many Ways to Begin

The Many Roads to Japan

Eureka!: Math Fun from Many Angles

Eureka!: Math Fun from Many Angles

Quicksilver: A Revolutionary Way to Lead the Many and the Few -- Beginning with YOU

Paul Mccartney Many Years From Now

We Few

Eureka!: Math Fun from Many Angles

A Few Quiet Days

A Few Good Men

How Few Remain

A Few Green Leaves

How Few Remain

And Now a Few Laughs from Our Sponsor with CDROM

And Now a Few Laughs from Our Sponsor with CDROM

How Few Remain

A Few Good Men

Few Are Chosen

A Few Good Men

Many Waters

Many waters

From Few to Many

From Few to Many

Many Miles to Go

Many Lives, Many Masters

From Bulk to Nano: The Many Sides of Magnetism

From Bulk to Nano: The Many Sides of Magnetism

The Many Faces of Philosophy: Reflections from Plato to Arendt

So Many Ways to Begin

So Many Ways to Begin

Introduction to many-body physics

So Many Ways to Begin

The Many Roads to Japan

Eureka!: Math Fun from Many Angles

Eureka!: Math Fun from Many Angles

Quicksilver: A Revolutionary Way to Lead the Many and the Few -- Beginning with YOU

Paul Mccartney Many Years From Now

We Few

Eureka!: Math Fun from Many Angles

A Few Quiet Days

A Few Good Men

How Few Remain

A Few Green Leaves

How Few Remain

And Now a Few Laughs from Our Sponsor with CDROM

And Now a Few Laughs from Our Sponsor with CDROM

How Few Remain

A Few Good Men

Few Are Chosen

A Few Good Men

Many Waters

Many waters

Recommend Documents